Download FLAMES User Manual

EUROPEAN SOUTHERN OBSERVATORY
Organisation Européene pour des Recherches Astronomiques dans l’Hémisphère Austral
Europäische Organisation für astronomische Forschung in der südlichen Hemisphäre
ESO - European Southern Observatory
Karl-Schwarzschild Str. 2, D-85748 Garching bei München
Very Large Telescope
Paranal Science Operations
FLAMES User Manual
Doc. No. VLT-MAN-ESO-13700-2994
Issue 80, Date 25/12/2006
Prepared
A. Kaufer, C. Melo
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Date
Approved
A. Kaufer
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Date
Released
Signature
Signature
O. Hainaut
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Date
Signature
FLAMES User Manual
VLT-MAN-ESO-13700-2994
This page was intentionally left blank
ii
FLAMES User Manual
VLT-MAN-ESO-13700-2994
iii
Change Record
Issue/Rev.
Date
Section/Parag. affected
Reason/Initiation/Documents/Remarks
0.5
0.6
0.7
1.0
1.1
15/09/01
Apr. 02
Sep. 02
21/03/03
21/07/03
all
all
all
all
all
draft
draft
CfP P71, appendix added
First Release (for P71)
Updates and Corrections
Results ARGUS commissioning
Release for P72
HR update, IFU geometry,
calibration times added
Minor changes for P73
New HR settings added
New HR setting added
Update for P74
1.12
24/10/03 some
1.13
1.14
1.15
1.16
15/01/04
01/03/04
04/03/04
18/06/04
1.17
2
07/07/04
25/11/04
2
79
79.1
80
11/03/05
01/09/06
25/12/06
27/02/07
some
Table 3.1
Table 3.1
Sects 1.4, 1.5, 2.2, 3.1, 3.3.1,
Sects 4.4, 5.1.1, 6.2, 7.3
Figs 2.6, Tables 3.1, 7.1, 7.3
Sect 1.4
Sect 1.2, 1.3.2, 2.4.5, 1.4, 1.6
1.7, 2.8.1, 4.3, 5.1.1
some
none
Sect. 29 added
none
Combined cfg λ comment
Update for P75
Figures
Version
Version
Version
quality improved
for Period 79 Phase I and II
for Period 79 Phase II
for Period 80 Phase I and II
FLAMES User Manual
VLT-MAN-ESO-13700-2994
This page was intentionally left (almost) blank
iv
FLAMES User Manual
VLT-MAN-ESO-13700-2994
v
Contents
1 Introduction
1
2 On the Contents of the FLAMES User Manual
1
3 Information available outside this Manual
2
4 Capabilities of the Facility
4.1 UVES – FIBRE mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2 GIRAFFE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3 FLAMES Observing Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3
3
5
5
5 Limitations and Caveats
7
6 FLAMES within the VLT Observatory
10
7 FLAMES Sample Observations and Calibrations
11
8 High-resolution Spectrographs at ESO La Silla
11
9 Bibliography
11
10 Glossary
12
11 Abbreviations and Acronyms
14
12 FLAMES Characteristics and Sub-Systems
14
13 Opto-mechanical Layout
14
14 Corrector
15
15 Fibre Positioner (OzPoz)
15.1 Positioner Performance Characteristics . . . . . . . . . . . . . . . . . . . . . .
16
18
16 Buttons and Fibre Systems
16.1 Magnetic Buttons . . . . .
16.2 UVES Fibres . . . . . . .
16.3 MEDUSA Fibres . . . . .
16.4 IFU Fibres . . . . . . . . .
16.4.1 IFU Orientation . .
16.5 ARGUS Fibres . . . . . .
.
.
.
.
.
.
19
19
19
23
23
24
25
17 GIRAFFE
17.1 Slit Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
25
25
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
FLAMES User Manual
17.2
17.3
17.4
17.5
17.6
17.7
VLT-MAN-ESO-13700-2994
Filters and the Filter Wheel . . . . . . . . .
Dioptric Spectrograph . . . . . . . . . . . .
Scientific CCD “Bruce” . . . . . . . . . . . .
Spectral Format and Efficiency . . . . . . .
GIRAFFE Setup Stability and Repeatability
GIRAFFE Calibration Units . . . . . . . . .
.
.
.
.
.
.
27
27
28
28
30
30
18 UVES – FIBRE mode
18.1 The RED Spectrograph Arm . . . . . . . . . . . . . . . . . . . . . . . . . . . .
18.2 Scientific CCD Mosaic “STING”+“NIGEL” . . . . . . . . . . . . . . . . . . .
18.3 Spectral Resolution and Overall Efficiency . . . . . . . . . . . . . . . . . . . .
31
32
32
34
19 FLAMES Features and Problems
19.0.1 Maximum reachable S/N ratio . . . . . . . . . . . . . . . . . . . . . . .
19.0.2 Enhanced Dark Current after a FIERA Start-up . . . . . . . . . . . . .
35
35
35
20 GIRAFFE Features and Problems
20.1 Contamination from Simultaneous Th-Ar Calibrations . . . . . . . . . . . . . .
20.2 In-focus Ghosts and Scattered Light . . . . . . . . . . . . . . . . . . . . . . . .
20.3 CCD Defects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
35
35
36
36
21 UVES Features and Problems
21.1 Fibre Overlap in the 520 nm Setup .
21.2 Fibre-to-fibre Contamination . . . . .
21.3 Spectral Gaps in the RED . . . . . .
21.4 Optical Ghosts in the far red Spectra
21.5 CCD Cosmetic Defects . . . . . . . .
36
36
36
37
38
38
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
vi
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
22 Preparing the Observations
38
23 Introduction
38
24 FLAMES Modes and basic Choices
40
25 GIRAFFE and UVES Standard Settings
25.1 GIRAFFE Standard Settings . . . . . . . . . . . . . . . . . . . . . . . . . . .
25.2 UVES Standard Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
41
41
43
26 Differential Atmospheric Effects
43
27 Preparing the Target Input Files
48
28 Run FPOSS to Prepare the Target Setup Files
49
29 Broken Fibers
50
FLAMES User Manual
VLT-MAN-ESO-13700-2994
vii
30 Introducing the Observation Blocks
51
31 GIRAFFE and UVES Exposure Time Calculators
31.1 Choice of the Sample Target . . . . . . . . . . . . . . . . . . . . . . . . . . . .
31.2 Choice of Instrument Configuration and Spectral Format . . . . . . . . . . . .
31.3 Exposure Time and predicted Counts and S/N Ratios . . . . . . . . . . . . . .
51
51
52
52
32 P2PP tool
32.1 Acquisition Templates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
32.2 Observing Templates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
32.3 Computing Time Overheads for your Programme . . . . . . . . . . . . . . . .
52
52
53
53
33 The Calibration of FLAMES Data
54
34 General Concept
54
35 Positioner Calibration Unit
54
36 Nasmyth Screen
55
37 Simultaneous Calibrations
55
38 Longslit Calibrations
38.1 GIRAFFE Longslit Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
38.2 The UVES Calibration Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . .
56
56
56
39 Fibre to Fibre Transmission (Sky Subtraction)
56
40 Special Calibrations
40.1 Detector Flats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
40.2 Use of Telluric Standard Stars to correct for Fringing or atmospheric Lines . .
57
57
57
41 FLAMES Science Calibration Plan
57
42 FLAMES Observing Operations
57
43 During the Night
43.1 Pointing and Guiding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
43.1.1 ARGUS fast observations . . . . . . . . . . . . . . . . . . . . . . . . .
43.2 Evaluation of the Results, Offline Data Analysis . . . . . . . . . . . . . . . . .
59
59
60
61
44 FLAMES Raw Data Structure
61
45 HDU2: OzPoz table
61
46 HDU3: FLAMES FIBRE Table
63
FLAMES User Manual
VLT-MAN-ESO-13700-2994
viii
47 Appendix
64
48 FLAMES Raw Data Spectral Format
48.1 GIRAFFE - MEDUSA . . . . . . . .
48.2 GIRAFFE - IFU . . . . . . . . . . .
48.3 GIRAFFE - ARGUS . . . . . . . . .
48.4 UVES - FIBRE . . . . . . . . . . . .
64
65
66
67
68
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
49 Characteristics of GIRAFFE Filters
69
50 FLAMES calibration times
74
51 Comparison between old and new HR gratings
75
List of Figures
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Schematic view of an Integral Field Unit . . . . . . . . . . .
MEDUSA entrance losses due to fibre-object decentering . .
View of the Fibre Positioner and GIRAFFE on the Nasmyth
TCCD image of the Fiducial Stars . . . . . . . . . . . . . . .
Histogram of Fibre Transmission at 600 nm . . . . . . . . . .
UVES Fibre Bundles and Slit Geometry . . . . . . . . . . .
Schematic view of Microlenses . . . . . . . . . . . . . . . . .
MEDUSA Fibre Bundles and Slit Geometry . . . . . . . . .
ARGUS Microlens Array and Slit Geometry . . . . . . . . .
GIRAFFE Spectral Format and Slit Curvature . . . . . . . .
UVES Fibre-to-Fibre Contamination . . . . . . . . . . . . .
Atmospheric Dispersion Effects at 9 arcmin from center . . .
Atmospheric Dispersion Effects at 12.5 arcmin from center .
Chromatic Atmospheric Dispersion Effects . . . . . . . . . .
FPOSS Sample Input File . . . . . . . . . . . . . . . . . . .
ARGUS reconstructed image . . . . . . . . . . . . . . . . . .
GIRAFFE - MEDUSA Spectral Format . . . . . . . . . . . .
GIRAFFE - IFU Spectral Format . . . . . . . . . . . . . . .
GIRAFFE - ARGUS Spectral Format . . . . . . . . . . . . .
UVES - FIBRE Spectral Format . . . . . . . . . . . . . . . .
GIRAFFE Filters HR 01-06 . . . . . . . . . . . . . . . . . .
GIRAFFE Filters HR 07-12 . . . . . . . . . . . . . . . . . .
GIRAFFE Filters HR 13-18 . . . . . . . . . . . . . . . . . .
GIRAFFE Filters HR 19-22, LR 01-02 . . . . . . . . . . . .
GIRAFFE Filters LR 03-08 . . . . . . . . . . . . . . . . . .
. . . . . . .
. . . . . . .
A Platform
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6
9
15
18
20
21
22
24
26
30
37
45
46
47
49
64
65
66
67
68
69
70
71
72
73
FLAMES User Manual
VLT-MAN-ESO-13700-2994
1
Introduction
2
On the Contents of the FLAMES User Manual
1
The current version of the FLAMES user manual is available online as a retrievable postscript/pdf
file at the ESO home page (http://www.eso.org/observing/vlt/instruments/flames/).
Prior to the observing proposal application and/or phase 2 announcements, the User Manual
is usually updated; any significant changes are announced on the FLAMES web pages. If
you do not have access to the WWW, a printed copy can be requested from ESO’s Visiting Astronomers Section (e-mail: [email protected]) in Garching, Germany. Paper copies of a
new version of the FLAMES User Manual are printed out only after a major revision of the
document.
The first Chapter of this manual is addressed to users who are not familiar with the FLAMES
facility and who are interested in a quick overview of its capabilities, as in the case of
similar VLT (and La Silla) instruments. This should enable a potential user to select the best
instrument for a given observing program. It also includes information on how to access FITS
files of reference FLAMES spectra and a glossary of terms used in the Manual.
The second Chapter provides the description of the instrument: the instrument layout
(§2.1), its main components (Corrector, §2.2; Fibre Positioner, §2.3, Fibre System, §2.4), the
properties of GIRAFFE and UVES (§2.5, 2.6) spectrographs, including their resolving power
and overall efficiency. In addition, it contains the requirements to be kept in mind while
planning the observations or reducing the data. It can be consulted by users who want to
prepare an Observing Proposal (Phase I), but should definitely be read by those who have
been granted observing time and have to prepare their observations (Phase II). In particular,
the description of the Atmospheric Effects affecting FLAMES observations and
their consequences on planning and optimizing the observations is of fundamental
importance.
The third Chapter presents the basic information needed to prepare an observing programme:the various observing modes (§3.2),the standard wavelength settings (§3.3), and a
FLAMES User Manual
VLT-MAN-ESO-13700-2994
2
description of the Exposure Time Calculator (§3.8). This chapter explains how to prepare a target input file and how to generate a positioner allocation file. It assumes
that the reader is familiar with the fibre assignment software (FPOSS) and with the FLAMES
templates. The FPOSS manual and template descriptions will be provided as separated documents before Phase II.
The fourth Chapter deals with the calibration strategy (wavelength, flat-fielding, relative
and absolute calibrations) of the data obtained in standard operation. It also outlines the
calibration techniques for high velocity accuracy and demanding sky subtraction. The fifth
Chapter provides information for the visiting astronomers who come to the Paranal Observatory to use FLAMES. The sixth Chapter summarizes the properties of the pipeline reduction, which will be applied to the data as they become available (cf. http://www.eso.org/qc/
for up-to-date information).
The detailed technical information on the instrument (e.g., transmission curves of the
GIRAFFE filters) can be found in the Appendices.
The FLAMES Templates Reference Guide [1] which contains detailed instructions for
the use of the observing and calibration templates, The FPOSS manual [2] which illustrates
the use of the positioner software for the allocations of the fibres to the objects are given as
separate documents, although they should be considered by the user as PART of the present
manual.
3
Information available outside this Manual
If you cannot find a specific piece of information in the FLAMES User Manual or in case you
have remaining questions, please contact http://www.eso.org/observing/support.html, or
more specifically:
• For information on the instrument performance, Phase I, and Phase II proposal preparation, please contact the User Support Division ([email protected]).
• For Phase II preparation of Service Mode Observation Blocks (OBs) follow the instructions given in the FLAMES-specific P2PP page
http://www.eso.org/observing/p2pp/FLAMES-P2PP.html
• For questions directly related to your granted observing run in Visitor Mode, please
contact Paranal Science Operations ([email protected]). Visitor mode specific information
on FLAMES is found at
http://www.eso.org/instruments/flames/vainfo.html.
• For updates on the instrument not yet recorded in the current version of the user manual,
consult the FLAMES web page at http://www.eso.org/instruments/flames/
• For technical information on the instrument not related to an observing programme, contact: optics and mechanics: Hans Dekker ([email protected]); the electronics systems:
Walter Nees ([email protected]); the CCD detector systems Roland Reiss ([email protected]);
the instrument software Peter Biereichel ([email protected]); fibres and fibre system
Gerardo Avila ([email protected]).
FLAMES User Manual
4
VLT-MAN-ESO-13700-2994
3
Capabilities of the Facility
FLAMES is the multi-object, intermediate and high resolution fibre facility of the VLT.
Mounted at the Nasmyth A platform of UT2 it offers a rather large corrected field of view (25
arcmin diameter) and it consists of several components:
• An optical Corrector, providing excellent image quality and tele-centricity over the full
field of view of 25 arcminutes diameter.
• A Fibre Positioner hosting two plates. While one plate is observing, the other one is
positioning the fibres for the subsequent observations, therefore limiting the dead time
between observations to less than 15 minutes.
• A link to the UVES spectrograph (RED arm) via eight single object fibres per plate.
• A high and intermediate resolution optical spectrograph, GIRAFFE, with its own fibre
systems in three possible configurations: MEDUSA, IFU, ARGUS.
• A coordinating observing software system, that allows simultaneous UVES and GIRAFFE observations.
The operation of FLAMES requires that the observer has her/his own target
coordinate list, with a relative astrometric accuracy better than ∼ 0.3 arcsec (rms)
at the time of the Phase 2 proposal preparation.
The minimum object separation is 11 arcsec, which is limited entirely by the size of the
magnetic buttons. The Fibre Positioner is able to position the fibres with an accuracy better
than ±0.1 arcsec (peak to peak).
In addition to the targets, the user must also provide coordinates for one VLT guide star and
four fiducial stars in the same astrometric solution as the targets. The VLT guide star is used
to first point the telescope and to close the active optics loop, while the four fiducials are used
to correct this pointing for further small offsets in coordinates due to corrections of the field
geometry.
4.1
UVES – FIBRE mode
UVES is the high resolution spectrograph of the VLT UT2. It has been designed for working
in slit mode only but was modified to add a fibre mode on its RED arm. Each positioner
plate has eight fibres connected to the red arm of UVES. In 520-nm mode only 6 of these are
available. With an aperture on the sky of 1 arcsec, the fibres project onto five UVES pixels in
the dispersion direction giving a resolving power of ∼ 47000.
In addition to the eight fibres per plate, an extra fibre fed by a separate calibration unit is
available. This fibre is used for simultaneous calibration in order to obtain very accurate radial
velocities. Only seven fibres can be devoted to astronomical objects when this simultaneous
calibration fibre is used. Note that this simultaneous calibration fibre mode is only available
in the 580 nm setup. For faint objects one or more fibres can be devoted to the sky.
When used in Fibre mode, only the three standard UVES RED setups are offered, with central
wavelength of 520, 580 and 860 nm, respectively (see the UVES user manual for details
http://www.eso.org/instruments/uves/).
FLAMES User Manual
Spectro.
UVES
UVES7
Mode
RED
RED
GIRAF
GIRAF
GIRAF
GIRAF
GIRAF
MEDUSA
MEDUSA
IFU
IFU
ARGUS
HR
LR
HR
LR
HR
GIRAF LR
ARGUS
Spectro.
UVES
UVES7
GIRAF HR
GIRAF LR
GIRAF HR
GIRAF LR
GIRAF HR
GIRAF LR
Mode
RED
RED
MEDUSA
MEDUSA
IFU
IFU
ARGUS
ARGUS
VLT-MAN-ESO-13700-2994
4
N. Objects
8 (with sky)
7 (with sky)
Aperture [00 ]
1.0
1.0
R
47000
47000
Cover.
200
200
131 a (with sky)
131 a (with sky)
15 (+15 sky)
15 (+15 sky)
1
1.2
1.2
2×3
2×3
11.5×7.3
or 6.6×4.2
11.5×7.3
or 6.6×4.2
V S/N 30
15.5
15.5
17.1
18.1
15.5**
16.5**
15.5**
16.5**
19000∗
7000∗
30000∗
11000∗
30000∗
λ/22 – λ/12
λ/9.5
λ/22 – λ/12
λ/9.5
λ/22 – λ/12
11000∗
λ/9.5
1
V S/N=10
17.5
17.5
18.7
19.7
17.5**
18.5**
17.5**
18.5**
”/pix
0.18
0.18
0.19
0.19
0.19
0.19
0.19
0.19
Comments
+Simul.
Calib.
RV accuracy ***
300 m/sec
30 m/sec
150 m/sec
300 m/sec
150 m/sec
300 m/sec
150 m/sec
300 m/sec
(∗): The resolving powers (R) given here are only average values, for details see Tables 3.1
and 3.2, which contain a description of all the GIRAFFE setups.
(∗∗) Magnitudes for IFU and ARGUS modes are given for extended objects, in surface brightness (magnitudes/arcsecond)
(∗ ∗ ∗) Radial velocity accuracy is estimated for a slowly rotating solar-like star over several
days. See Sect. 37. The on going analysis of commissioning data will provide better long-term
estimates.
(a) The number of allocatable buttons is 132, but only 131 spectra are fully covered on the
detector.
Table 1: FLAMES characteristics and observing capabilities. The wavelength coverage
(Cover.) is given is nanometers. The S/N ratio is given per wavelength pixel (as in the
ETC) and it refers to the mean S/N ratio in the setups LR4 (543.1 nm) and HR10 (548.8 nm),
using as inputs a G2 star for point-like and elliptical galaxy for extended sources. Additional
assumptions include 1 hour exposure, dark time, 0.8 arcsecond seeing, airmass 1.2 and a fibre
perfectly centered on the object.
FLAMES User Manual
4.2
VLT-MAN-ESO-13700-2994
5
GIRAFFE
GIRAFFE is a medium-high resolution spectrograph (R = 7500 − 45000) for the entire visible
range (370 − 950 nm). It is equipped with two gratings (high and low resolution) and uses
order sorting filters to select the required spectral range. Each object can be only observed in
one, or a fraction of a single, echelle order at once. GIRAFFE is equipped with a 2k×4k EEV
CCD (15 µm pixels), with a scale of 0.19 arcsec/pixel.
The fibre system feeding GIRAFFE consists of the following components:
• 2 MEDUSA fibre slits, one per positioner plate. Up to 131 different objects (including
sky fibres) are accessible in MEDUSA single fibre mode, each with an aperture of 1.2
arcsec on the sky. 5 additional fibres allow simultaneous calibration of every exposure.
• 2 IFU slits, one per positioner plate. Each deployable Integral Field Unit (IFU) consists
of an array of 20 square microlenses of 0.52 arcsec side each, giving a total (almost
rectangular) aperture of ∼ 3 × 2 arcsec. For each plate there are 15 IFUs dedicated to
objects and 15 IFUs dedicated to sky measurements; the latter contain the central fibre
only. A schematic representation of an IFU in its button is given in Figure 1.
• 1 ARGUS slit. The large integral unit ARGUS consists of a rectangular array of 22 by
14 microlenses. It is fixed at the center of positioner plate 2. Two scales are available:
one with a sampling of 0.52 arcsec/microlens (scale 1:1) and a total aperture of 12 by 7
arcseconds, and one with a sampling of 0.3 arcsec/microlens and a total coverage of 6.6
by 4.2 arcseconds (scale 1:1.67). 15 ARGUS single sky fibres are also available. These
can be positioned within the 25 arcmin field (cf. Figure 9). The ARGUS long axis
is along the N–S direction for a position angle of 0 degrees, with the PA
entered in FPOSS being measured North–East.
GIRAFFE is operated with 32 fixed setups (24 high resolution + 8 low resolution) whose
characteristics are given in Table 10 and Table 11.
For performance estimates (based on measured transmission curves and performances) the user
is referred to the Exposure Time Calculator (http://www.eso.org/observing/etc.html).
A summary of the GIRAFFE characteristics is given in Table 1.1, including estimated best
performances S/N ratios.
4.3
FLAMES Observing Modes
The FLAMES observing software (OS) coordinates the various observing modes (MEDUSA,
IFU, ARGUS). In addition, it allows a simultaneous acquisition of UVES and GIRAFFE
observations with the specific observing modes listed in Table 2. It is important to note that
during a combined observation the exposure times for UVES and GIRAFFE do not need to be
the same, but the longest exposure time will determine the overall length of the observation.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
6
Figure 1: Schematic representation of a deployable Integral Field Unit (IFU), in its button.
The signal from the rectangular microlens system (0.52 arcsecond squared per microlens) is
brought to the Giraffe spectrograph through 20 fibres. The fibres of one IFU form one subslit
of the IFU slit.
FLAMES User Manual
5
VLT-MAN-ESO-13700-2994
7
Limitations and Caveats
FLAMES is a complex instrument because of the different modes available and the multiobject capability. In order to operate it efficiently, a number of limitations had to be imposed.
• The observer is responsible for the accuracy of the input catalogue. A relative
accuracy of better than 0.3 arcseconds (rms) is required to limit the losses
due to fibre-object mismatch. VLT guide star and fiducial stars must have
coordinates in the same reference system as the objects.
No cross check of the coordinates supplied by the user is performed by ESO. The quality
of the astrometry remains fully the observer’s responsibility. Common errors include
using a mix of astrometric systems and/or not correcting for stellar proper
motions.
Figure 1.2 shows the amount of flux lost in a MEDUSA fibre as a function of seeing and
fibre-to-object decentering (in fraction of arcseconds); it is evident as bad coordinates
may spoil completely the predicted performance. The reader should consider the full
implications of the statistical meaning of the astrometric accuracy; if this is too bad,
some of the objects will not get light at all!. This factor is even more important for
the UVES fibres which are 1.0 arcseconds in diameter as opposed to 1.2 arcseconds for
MEDUSA fibres.
• Given the high number of possible configurations, the spectral format is fixed for both
GIRAFFE and UVES: no CCD binning, no CCD readout speed, no tuning of the wavelength, no change in resolving power are possible.
• Since the day-time calibration procedure is rather long (up to several minutes/setup,
especially in the bluest setups) only a limited number of setups may be allowed per
night, both in service and visitor modes.
• The atmospheric effects depend on the wavelength of observation. The VLT pointing
and guiding is made for a given wavelength. While it is possible in P2PP to specify
Spectrograph
Mode
Single Modes:
UVES
(a): 8 target fibres (580 nm or 860 nm setups)
UVES
(b): 7 target fibres + 1 calibration fibre
(580 nm setup only)
UVES
(c): 6 target fibres (520 nm setup)
GIRAFFE
MEDUSA
GIRAFFE
IFU
GIRAFFE
ARGUS
Combined Modes:
UVES + GIRAFFE UVES (a) and (b) + MEDUSA
UVES + GIRAFFE UVES (a) and (b) + IFU
UVES + GIRAFFE UVES (a) and (b) + ARGUS
Table 2: Summary of the various single and combined modes of FLAMES
FLAMES User Manual
VLT-MAN-ESO-13700-2994
8
two different observing wavelengths for the UVES and GIRAFFE fibres, the VLT will
point and guide only to the GIRAFFE wavelength in this combined mode.
This implies that the pointing will be correct, but if the airmass is changing drastically
during the observations and the UVES and GIRAFFE wavelengths differ dramatically,
then the UVES λ is disregarded and the UVES fibres may loose a considerable amount
of light. For the same reason, it is important that if the observer wishes to observe
the same objects at different wavelengths, two OBs (and therefore two different fibre
positionings) are made, repositioning the fibres for the correct wavelength. We finally
note that in a combined observation, if the UVES part is more important,
then the GIRAFFE configuration wavelength can be chosen to be close to the
UVES wavelength e.g. H572.8 in the case of using UVES-580 nm. Of course,
this would lead to entrance losses in the GIRAFFE part if the GIRAFFE
observation wavelength is different from the above.
• In general, long (e.g. longer than 60 minutes for most declinations, see Section 3.4)
and/or repeated observations of the same objects are better split in several observations
made with different plates. Since the geometry of the field will slightly change with time,
it is anyway recommended to reposition the fibres after each observation.
• The positioning time is about 10 seconds/fibre, or 20 minutes for MEDUSA. This implies
that OBs shorter than 20 minutes will suffer considerable deadtime before the next
observation is started. In these cases, the duty cycle is very bad; another instrument is
perhaps more suited for the observations. Note that, due to the implementation of the
UVES slit, the positioning of the UVES fibres for the next observation cannot be done
simultaneously with the current observations, but only after that the UVES shutter has
been closed. Repositioning of UVES fibres requires 90 seconds in total.
• The UVES simultaneous wavelength Th-Ar lamp can accept exposure times in the range
of 3 − 60 minutes for the 580 nm setup giving an acceptable level of exposure. Shorter or
longer exposure times, however, will result in under (and over) exposed Th-Ar reference
spectra, respectively. These will not be accepted in Service Mode. In the 860 nm setup
long exposures would provide heavily saturated Ar lines, producing strong, persistent
remnants. Neither the 860 or the 520 nm setups are offered with the simultaneous
calibration option.
• All observations must be prepared with the FPOSS preparatory tool (See chapters 3.5
and 3.6 and the FPOSS manual). The Target Setup Files created by this tool must not
be modified by the user. This will cause the P2PP process to fail.
• The limited size of the MEDUSA and UVES fibres, together with the lack of information
on the object-fibre displacement makes it impossible to compute the amount of flux lost;
therefore no absolute spectro-photometry can be obtained with these fibre systems.
Unlike other multi-object ESO instruments, FLAMES does not have pre-imaging capabilities
to prepare target selection. This implies that astrometric lists must be prepared by the
observer. Experience with other similar instruments shows that most observation failures are
due to improper target preparation. Also, given the relatively large field, atmospheric effects
(e.g., differential refraction and its variations, see Section 3.4) may be relevant, and the reader
is asked to consider them carefully when preparing the observations.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
9
Figure 2: MEDUSA entrance losses as a function of seeing and object-fibre decentering. This
plot shows how much flux can be lost due to bad astrometry. The reader should evaluate the
impact of the astrometric errors in their full statistical sense.
FLAMES User Manual
6
VLT-MAN-ESO-13700-2994
10
FLAMES within the VLT Observatory
A detailed overview of the different instruments on the VLT is given on the ESO homepage
under VLT Instrumentation (http://www.eso.org/observing/vlt/instruments/). In the
choice of the best instrument for a given observing programme, the following tradeoffs have
to be considered:
VLT instruments that can perform spectroscopy in the UV-Visual-Red-regions
(300 − 1100 nm)
• FORS1 at UT2 can be used for spectroscopy in the spectral range 360 − 1100 nm. Its
overall efficiency is on average 3-4 times higher than GIRAFFE and UVES-Fibres (one
reflection less in the telescope and simpler instrument optics) but the maximum resolving
power to be obtained with a 0.5 arcsec slit is 2500 only. As far as multi-object spectroscopy (MOS) is concerned, a multiplex of 19 is achieved on a field of view of 6.8 by 6.7
arcminutes. See the FORS webpage for details at http://www.eso.org/instruments/fors/.
• FORS2 at UT1 is a replica of FORS1, optimized for the RED part of the spectrum. In addition to the MOS capabilites of FORS1 masks with up to 200 slitlets
can be inserted in the same field of view of FORS1. The highest resolution possible
is ∼ 6000, although only with certain setups. See the FORS webpage for details at
http://www.eso.org/instruments/fors/.
• UVES at UT2 is the instrument which is closest to FLAMES in terms of spectral
resolution. In slit mode, the resolving power of UVES can be up to 120 000. The UVES
red arm is also part of FLAMES, but its blue arm (300 − 500 nm) is not connected to
FLAMES. When used in slit mode, with a dichroic blue and red spectra can be recorded
simultaneously. This option is not available in UVES fibre mode with FLAMES.
When used in slit mode, the RED arm of UVES is about 2 times more efficient than
the FLAMES fibre link. The multiplex advantage of using the FLAMES fibre link with
respect to the slit mode can therefore be ≈ 3 − 4 (depending if one or more fibres are
dedicated to record the sky). This has to be considered just as a rough number, because
the precise value will depend on the seeing and on resolving power adopted for UVES
in slit mode.
GIRAFFE, in particular in IFU mode, can approach the typical resolution used with
UVES, and their use can be considered as a valid alternative to UVES slit when several
sources are present in the field and a very large wavelength coverage is not required.
UVES is equipped with a Iodine cell for accurate radial velocity measurements. While
this system is likely more accurate than the multi-fibre system of FLAMES, it does not
offer multiplex capabilities. The UVES iodine cell cannot be used in combination with
the fibres because it is located in the focal plane of Nasmyth B, i.e., before the Fibre
link to FLAMES.
• VIMOS has a smaller field of view than FLAMES (a square of 14 × 14 arcminutes), but
a higher multiplex gain (up to 400 mini-slits punched in to mask). The major difference
is the spectral format and a lower resolution (R = 4500 for a 0.5 arcseconds wide slit).
VIMOS also has a Integral Field Unit. This Unit is larger than the GIRAFFE-ARGUS:
it may be as large as 6000 by 6000 ) with a resolving power of 300 or as large as 30” by 30
” with a resolving power of 2000.
FLAMES User Manual
7
VLT-MAN-ESO-13700-2994
11
FLAMES Sample Observations and Calibrations
A large number of scientific observations of a variety of targets and their associated calibrations have been obtained during the FLAMES Commissioning and Science Verification runs.
They have been made publicly available at http://www.eso.org/science/flames comm and
http://www.eso.org/science/vltsv/flamessv. FLAMES calibrations are available from
the ESO archive at http://archive.eso.org/.
8
High-resolution Spectrographs at ESO La Silla
Other high-dispersion, echelle type spectrographs available at ESO La Silla include FEROS
at the 2.2-m telescope, HARPS at the 3.6-m telescope and EMMI at the NTT. (Very) high
dispersion (up to a resolving power of 235,000) in a single spectral order is provided by the
CES Very Long Camera, which is fed by the 3.6-m telescope via a fibrelink. HARPS has been
recently commissioned at the 3.6-m telescope and is a fibre-fed echelle spectrograph dedicated
to most accurate radial-velocity measurements.
9
Bibliography
[1 ] FLAMES Templates Reference Guide, VLT-INS-MAN-ITA-13750-0009, version 1.3, E. Rossetti et al.
[2 ] FPOSS User Manual, VLT-INS-MAN-AUS-13271-0079, version 1.9, K. Shortridge et al.
[3 ] FLAMES Calibration Plan, VLT-PLA-ESO-13700-xxx, Issue 0.3, 21/07/03, A. Kaufer
[4 ] Data Interface Control Document, GEN-SPE-ESO-19400-794, Issue 1.1, 25/11/97, M.
Albrecht
[5 ] P2PP Users’ Manual, VLT-MAN-ESO-19200-1644, version 2.2,17/1/2002, D. Silva
[6 ] Interface Control Document between the VLT Control Software and the Observation Handling Subsystem, VLT-ICD-ESO-17240-19200, Issue 1.1, 17/10/97, E. Allaert, A.M. Chavan
[7 ] CCD DCS Dictionary, ESO-VLT-DIC.CCDDCS, Version 2.12, 17/04/1998
[8 ] FIERA DCS Dictionary, ESO-VLT-DIC.FCDDCS, Version 2.25, 02/04/1998
[9 ] TCS Dictionary, ESO-VLT-DIC.TCS, Version 1.66, 14/04/1998
[10 ] FLAMES papers presented at SPIE 2000 and 2002: To be added (see also FLAMES
web page)
[11 ] UVES User Manual http://www.eso.org/instruments/uves/userman/
FLAMES User Manual
10
VLT-MAN-ESO-13700-2994
12
Glossary
Acquisition: Accurate positioning of the telescope in order to center the target on the
spectrograph slit.
BIAS frame: Read–out of the CCD detector of a zero seconds integration time exposure
with shutter closed. The registered number of electrons per pixel has to be subtracted
from a science exposure, because these were not created by photons from the source.
Calibration: Procedures to remove the instrumental signature from the scientific data (e.g.,
subtract BIAS frames and divide by the flatfield).
Camera: GIRAFFE and UVES have dioptric cameras imaging the dispersed parallel beams
on the respective CCD detectors.
Charge-Coupled Device (CCD): Electronic 2D-array detector converting photons into
electrons.
Cross-disperser grating: An echelle spectrograph contains two dispersive elements. One
is the echelle grating, the other one is called the cross-disperser grating. UVES hosts two
cross-dispersers, each with two different gratings. The cross-disperser grating determines
the distance between the echelle orders.
Decker: Reflecting and movable blades placed in front of the slit and determining its length.
FACB: Fiducial Acquisition Coherent Bundles. These are 4 bundles of coherent fibres to
take images of 4 reference stars and link the telescope to the targets.
Flatfield (FF): Spectrum obtained from light source with a flat (i.e. without spectral features) energy distribution, e.g. a tungsten lamp. The registered signal provides information about the response of the detector, allowing a determination of the variation in
sensitivity from pixel to pixel, the echelle order shape, the presence of bad columns on
the detector, etc.
Grating: The main light dispersing elements of UVES and GIRAFFE are echelle gratings.
Guide star: A point source used for accurate tracking (and active control of the telescope
mirrors) magnitude between R∼11 and R∼13.
Maintenance: Technical procedures developed to control and maintain the quality of telescope, instrument, and detector.
Observation Block (OB): A logical unit of exposures needed to obtain a coherent set
of data. Encompasses all relevant information for a successful data acquisition on a
target. It consists of target information, a set of templates, parameter files for the
templates, conditions, requirements and comments concerning the specified observations.
It represents the entity the short-term scheduler deals with. Constructing Observation
Blocks is part of the Phase II Proposal Preparation Process.
Order Separation Filters: In GIRAFFE the wavelength range covered in each setup is
defined by using filters as predisperser: inserted in the beam they reject all the light
outside the defined bandpass, which instead is dispersed by the echelle grating.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
13
Phase II Proposal Preparation (P2PP): During this phase the successful applicant (whose
Phase I proposal has been accepted based on the scientific rationale and technical feasibility) prepares the Observation Blocks to carry out the observing programme.
Focal Plates: The Fibre Positioner can host up to 4 plates: these are metallic spherical
surfaces where the fibre buttons are positioned for the observations. Only two plates are
only currently in use.
Pre-slit area: UVES optical elements located in front of the spectrograph slits.
Spectrograph arm: UVES consists of two “separate” spectrographs, one optimized for the
blue (blue arm) and one for the red wavelength region (red arm). Only the red arm is
connected to the fibres
Spectrograph slit: Two parallel, reflecting metal blades with an adjustable separation (slit
width) form the entrance slit of the spectrograph. The image of the astronomical source
produced by the telescope is focused on the slit plane.
Standard Setting: A pre-defined setting of the instrument facilitating the preparation of
the observations. The Observatory keeps an updated database of the relevant calibration
files for all Standard Settings of the instrument.
Template: A set of instructions for the performance of a standard operation on an instrument, typically an instrument and detector setups. The templates represent specially
devised sequences for all instrument operations and calibrations.
Template Signature File: This file is a description of a Template and its parameters. It
contains information about the type and allowed ranges of the parameters; some of the
parameters have to be set by the observer.
Wavelength calibration: Spectrum obtained from a reference emission-line lamp (e.g. ThAr). The wavelengths of the (many) emission lines are accurately known and are used
to transform pixel space into wavelength space.
FLAMES User Manual
11
VLT-MAN-ESO-13700-2994
14
Abbreviations and Acronyms
AT
BOB
CAL
CCD
CD
ESO
ETC
FLAMES
FPOSS
FRD
IFU
OB
OS
P2PP
RTD
STD
SM
TSF
UVES
VLT
VM
Acquisition Template
Broker for Observation Blocks
Calibration exposure
Charge-Coupled Device
Cross-disperser
European Southern Observatory
Exposure Time Calculator
Fibre Large Array Multi Element Spectrograph
Fibre Positioner Observing Support Software
Focal Ratio Degradation
(deployable) Integral Field Unit
Observation Block
Observation Software
Phase II Proposal Preparation
Real Time Display
Standard star
Service Mode
Template Signature File
Ultraviolet and Visual Echelle Spectrograph
Very Large telescope
Visitor Mode
12
FLAMES Characteristics and Sub-Systems
13
Opto-mechanical Layout
Figure 3 is a view of two of the main components of the FLAMES facility, the Fibre Positioner
and GIRAFFE, as seen from the telescope centerpiece on the telescope platform.
The instrument consists of five main parts. The first part is the corrector which is mounted
on the rotator. The second part is the fibre positioner which allocates the fibres on the two
plates mated to the Nasmyth adaptor-rotator. The positioner also hosts the calibration lamps
used to obtain flat-field and wavelength calibration spectra. Furthermore, it is equipped with
a secondary astrometric and guiding system (FACBs) which consists of four imaging fibre
bundles correcting small mismatches between the VLT and the observer coordinate system.
These first two components are common to all FLAMES configurations.
The light is collected through fibres equipped with microlenses into different fibre systems:
two for UVES (one per plate) and five for the GIRAFFE spectrograph (two for MEDUSA,
two for IFUs, and one for ARGUS). The different fibres have different diameters and lengths
and are organized in different slit systems, each feeding the spectrographs.
Finally the light reaches the last two components: the UVES (RED) and the GIRAFFE
spectrographs, where it is dispersed and detected.
The next sections describe the FLAMES subsystems as one follows the optical path going from
FLAMES User Manual
VLT-MAN-ESO-13700-2994
15
Figure 3: The Fibre Positioner and GIRAFFE as seen during the GIRAFFE integration on
the Nasmyth platform. The picture is taken from the telescope. The positioner is looking
towards the Nasmyth Focus, where the corrector is placed, and on the lower left the positioner
electronics cabinet is seen. GIRAFFE is opened, and the optomechanical components are
visible.
the telescope to the instrument detectors. It is intended to guide the user in the selection of
the optimal instrument configuration for the observing programme. The functionalities of the
different subunits are explained and reference is made to their measured performance.
Efficiencies (e.g., in the form of transmission curves) of the main instrument components
including the CCDs are available in the FLAMES database accessible through the GIRAFFE
and UVES Exposure Time Calculators (see Chapter 3.5).
14
Corrector
The optical corrector is a doublet of BK7 equivalent lenses of 900 mm diameter. In order to
maintain a good transmission over a large wavelength range, the lenses have been coated with
a single layer of Mg2 .
The function of the corrector is to give an excellent image quality over the whole 25 arcmin
FLAMES field of view and to provide a pupil located at the center of curvature of the focal
FLAMES User Manual
Wave
[nm]
365
405
486
586
656
800
1014
0
0.789
0.894
0.919
0.914
0.898
0.880
0.843
VLT-MAN-ESO-13700-2994
Distance from optical axis [arcmin]
2
4
6
8
10
11
0.788 0.788 0.784 0.776 0.754 0.735
0.890 0.888 0.888 0.881 0.842 0.848
0.912 0.906 0.902 0.903 0.900 0.890
0.905 0.897 0.890 0.889 0.895 0.887
0.887 0.879 0.871 0.869 0.879 0.872
0.869 0.858 0.849 0.846 0.854 0.854
0.830 0.819 0.809 0.805 0.811 0.816
16
12
0.692
0.802
0.847
0.856
0.842
0.826
0.792
Table 3: Full Corrector transmission as function of wavelength and radial distance of the
object from center (in arcminutes). It includes as well pupil decentering effects for a MEDUSA
aperture.
plate.
The corrector is mounted with a cross support onto the Nasmyth adaptor-rotator. The support
also hosts the three attaching points for the Positioner fibre plates.
When the whole optical train is taken into account (including telescope optics and vignetting),
the effective transmission of the corrector depends on the observing wavelength and on the
distance of the object to the field center, as expressed in Table 3.
The corrector, and therefore the FLAMES plates, are positioned in the optical path AFTER
the VLT guide probe. This implies that the guide probe will vignet the field of view. It is
therefore very important to select carefully the VLT guide star. VLT guide star should have
an R magnitude between R∼11 and R∼13 for optimal performance. Although fainter objects
(to R<14) may work, experience has shown that due to uncertainties in the magnitude and
non-ideal observing conditions (e.g. cirrus or poor seeing), the Active Optics loop may fail to
close. If this occurs, then another guide star would need to be chosen that would likely vignet
the fibres on the plate.
Finally, it is also very important that the guide star is sufficiently isolated to avoid confusion
in its choice.
15
Fibre Positioner (OzPoz)
The Fibre Positioner (“OzPoz”) is at the core of the FLAMES facility. OzPoz is a rather
large and complex system equipped with four plates, two of which are currently in use (see
Figure 3). The Positioner can be subdivided into the following subsystems:
• Plates: Two metallic dishes, on which the magnetic buttons holding the fibres are attached. Each of the plates has a hole in the center. In one plate this hole hosts ARGUS.
Each plate has a curvature of 3950 mm, to match the curvature of the corrector focal
plane. The corrector also places the telescope exit pupil at the center of curvature of the
plate, so fibres receive the full telescope beam regardless of their position on the plate.
• Retractors: Mechanical systems maintaining the fibres in constant traction. Each fibre
is equipped with one retractor. The retractors are the same for all fibres. When parked,
FLAMES User Manual
VLT-MAN-ESO-13700-2994
17
the fibres are deposited and left on a porch located just outside the plate. The disposition
of the fibres on the plate(s) is similar, in that MEDUSA, UVES, FACB, IFU, and IFU
sky retractors are disposed in the same way on the two plates; every even-numbered
retractor is a MEDUSA one.
• Trolley: Main structure holding the plates. The trolley can perform two main movements: it can approach (or retract) the Nasmyth adaptor-rotator to engage the plate
(or disengage it). Furthermore it can rotate the structure holding the plates in order to
exchange the plates between the adaptor-rotator and the positioner robot.
• R-θ system and gripper (“robot”): This unit is at the very core of the whole system.
It grips and releases the magnetic buttons at the positions reached via the R-θ (polar)
robot. The gripper requires a back-illumination system, that means some light shining
through the fibres from the spectrograph towards the plate. A video system records this
back-illumination light and performs an image analysis for two purposes: first to reach
the required high position accuracy of the optical center of the fibre button and second
to detect if the magnetic button is properly picked by the gripper and properly released
on the plate. The polar coordinates of a placed fibre are stored in an internal permanent
memory (“NVRAM”) and are used for the next positioning.
• OzPoz is equipped with a calibration box, which moves with the gripper. This calibration
box hosts an optical system which directs the light either from a tungsten lamp, or
from a Th-Ar lamp, or from a Ne lamp into a fibre. In this way FF, Th-Ar and Ne
calibrations can be obtained for GIRAFFE and for UVES. The procedure to acquire
these calibrations is to first position the fibres to be calibrated on the plate in a given
pattern (typically along a spiral pattern) and then to sweep with the gripper over the
buttons, illuminating one fibre after the other, one by one. For FF calibration the
procedure is to sweep continuously over the buttons, illuminating them several times,
while for the Th-Ar calibrations the gripper stops over each fibre for a number of seconds
specified by the user.
• Field Acquisition Coherent Bundles (FACBs): Four magnetic buttons are equipped with
a system of 19 coherent fibres each. This bundle of fibres is used to obtain images of
“fiducial” (or reference) stars, one per bundle. The four images are recorded on an
ESO technical CCD (TCCD); the image centroids are computed and the proper offsets
are calculated to center the fiducial stars into the bundles. These 4 fiducial stars
represent the link between the sky and the plate coordinates, therefore it is
absolutely necessary that they are chosen carefully: They must be sufficiently
isolated, in the same coordinate system as the target stars and of visual
magnitude brighter than R = 15. Given the limited dynamical range of the
Technical CCD, the FACB stars should be within a range of 3 magnitudes.
Each FACB bundle has an effective diameter of 2.400 . One example of a TCCD image
with the four stars in the FACB bundles is shown in Figure 4.
• Positioning Software: This is based on a well-tested and complex code developed initially
for the 2dF system at Anglo-Australian Observatory (AAO). This so-called “delta”task allows crossing of the fibres in their final positions and determines the button
movements sequence from a given fibre configuration to the next (solving a traveling
salesman problem). It is worth noticing that to reach a new complex configuration more
FLAMES User Manual
VLT-MAN-ESO-13700-2994
18
Figure 4: Image of Four fiducial stars as seen in the Technical CCD through the 4 Fiducial
Acquisition Bundles (FACBs). Each bundle is composed of 19 ordered fibres, which image the
star on 4 quadrants of the TCCD.
than one move per fibre may be needed. A more detailed explanation can be found in
the FPOSS manual.
15.1
Positioner Performance Characteristics
The main performance characteristics of the Fibre Positioner can be summarized as follows.
1. Mechanical Positioning accuracy: better than 0.08 arcsec.
2. Positioning time: 6 seconds/move. The whole configuring time also depends on the
number of moves necessary to re-configure the plate, that is, the number of moves per
fibre needed to re-position one button, as well as by the number of attempts per button
needed to achieve the required positional accuracy.
3. Possibility to configure the next observation while observing.
4. Plate Exchange time: less than 180 seconds (w/o considering field acquisition).
5. Calibration unit equipped with Th-Ar, Ne and FF lamps
FLAMES User Manual
VLT-MAN-ESO-13700-2994
19
6. Performing fine centering of the plate on the sky with FACBs.
7. Minimum button separation 11 arcseconds (button diameter: 10 arcsecs).
16
Buttons and Fibre Systems
FLAMES is equipped with different types of fibres for UVES and for the different modes
of GIRAFFE. At the output of the fibre system, individual fibres are arranged in different
subslit systems depending on the fibre type. Each GIRAFFE mode has five fibres pre slit
devoted to simultaneous wavelength calibration, in addition to the fibres coming from
the Positioner. These fibres provide five calibration spectra for each observation acquired with
GIRAFFE. The UVES system has a similar simultaneous calibration capability: in that case,
one of the eight fibres is reserved for calibration.
In the evaluation of the instrument performance it has to be considered that among such a
large number of fibres some dispersion exists in the fibre transmission. Fibre transmission
within the GIRAFFE F/5 and UVES F/10 apertures have been measured for every single
fibre, and the distribution of the transmission is given in Figure 2.3 for the different fibre
types.
16.1
Magnetic Buttons
The Magnetic Buttons have two purposes: first, they are the mechanical means which allows
the fibre end to be placed on the plate in front of the stellar target. Second, they are the
support of the optical interface between the telescope beam and the fibre. A magnet glued
at the base of the button enables the fibres to be placed on the curved plate. The diameter
of the stainless steel button is 6 mm (10.3 arcsec), but in order to avoid collisions between the
buttons, the minimum allowed separation of two buttons is set to 11 arcsec. The magnet is a
tablet of 4 mm diameter and 1.2 mm high. The magnetic attachment force is around 235 gr.
Single microlenses for MEDUSA and UVES fibres and arrays of microlenses for ARGUS and
the IFUs are used to reduce the F/15 telescope beam to around F/5 into the fibres. These
lenses image the pupil of the telescope onto the fibre entranced surfaces. This system scrambles
the photometric variations produced by oscillations of the star on the fibre by guiding errors.
In the case of MEDUSA and UVES fibres, the optics are rod lenses with their image focal
plane on the flat surface. The lens has been cut to 45 degrees to fold the telescope beam into
the fibre, parallel to the plate. In the case of ARGUS and IFUs, all arrays of microlenses are
glued to a single total reflection 45 degrees prism.
16.2
UVES Fibres
Each of the positioner plates hosts eight 54-meters long fibres which bring the light to the
UVES spectrograph on the Nasmyth platform B.
The UVES fibre concept is shown in figure 6: there are two bundles (one per plate), each with
eight buttons. Every button hosts one fibre. From the UVES simultaneous calibration box,
one additional 5-meter fibre reaches the UVES-fibre slit.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
20
Figure 5: Distribution of the transmission of the FLAMES fibres at 600 nm. each fibre has
been measured in laboratory.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
21
Figure 6: Scheme showing the buttons, fibre bundles, and the geometry of the UVES slit.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
22
Figure 7: Scheme of the different microlens concepts.
Each of the two UVES-fibre slits (one per plate) consists of 9 fibres, although, only eight
fibres can be used simultaneously. They have a core of 120 microns diameter and a
cladding of 144 microns. Each fibre is protected by a Polyamide jacket of 180 micron diameter.
The fibre to fibre separation (center to center) is 1.7 times the fibre core (1.7 arcseconds).
The UVES system works at the optimal F/3 focal ratio, to minimize the focal ratio degradation
(FRD) and therefore the transmission losses. The exit is also at F/3; each fibre has an aperture
of 1 arcsecond on the sky. A microlens in front of the fibre converts the F/15 focal ratio of the
VLT Nasmyth focus to F/3; behind the microlens the light is reflected towards the side of the
button, where the entrance of the fibre is located (see Figure 2.5). The overall transmission
of the UVES fibre system is given in Table 2.3.
The fibres of each plate are arranged in one subslit (see Figure 6). The nine fibre centers are
separated by 1.7 times the fibre core diameter, implying that there is some degree of contamination between adjacent fibres. This contamination can be largely reduced by extracting the
spectra on the central six or seven pixels. Diffuse light is present, and since this depends on
the overall light injected into the spectrograph, the observer should be careful not to expose
FLAMES User Manual
Fibre
UVES
VLT-MAN-ESO-13700-2994
23
370 nm 400 nm 450 nm 600 nm
0.36
0.41
0.52
0.61
Table 4: UVES fibre transmission. The values given here include all losses, focal ratio degradation, optics and coupling. For wavelengths redder than 600 nm the transmission is constant.
Variations of a few tens percents between different fibres have been measured.
Fibre
MEDUSA
ARGUS
IFUs
370 nm 400 nm 450 nm 600 nm
0.47
0.52
0.55
0.61
0.52
0.58
0.62
0.70
0.49
0.55
0.58
0.66
Table 5: GIRAFFE Fibre Transmission: The values given here include all losses, focal ratio
degradation, optics and coupling. For wavelengths redder than 600 nm the transmission is
constant. Variations of a few tens percent between different fibres are measured.
objects of too different spectral type and/or luminosity if absolute spectroscopy is desired.
The simultaneous calibration fibre is the last one on the slit (Figure 2.4). The two subslit
centers are separated by 500 microns, (see Figure 2.4), therefore spectra obtained with different
plates are shifted by ∼40 pixels in the spectral direction on the CCD.
16.3
MEDUSA Fibres
Each plate also hosts 132 MEDUSA fibres. Each button includes a single fibre and its construction is similar to that of UVES. In the case of MEDUSA, the used focal ratio is F/5. The
core of each MEDUSA fibre is 230 microns, which corresponds to an aperture on the sky of
1.2 arcsecs. They have a cladding of 253 microns and a protection buffer of 280 microns. The
MEDUSA fibres are 13 meters long and their typical overall transmission is given in Table 2.3.
They are organized in a slit composed of several subslits. The MEDUSA subslits are of two
types: a) eight subslits hosting nine object-fibres and b) five subslits hosting thirteen (twelve
object + one simultaneous calibration) fibres. This fibre slit follows the curvature of focal
plane of GIRAFFE. The center-to-center distance of the MEDUSA fibres is of 2.26 times the
fibre core diameter: this ensures a fibre-to-fibre contamination below 0.5%.
16.4
IFU Fibres
Each Integral Field Unit (IFU) button is composed of twenty microlenses arranged in a rectangular shape (see Figure 1.1). The microlenses are 0.5200 squares. They convert the F/15
beam of the VLT to an F/7 focal ratio. Due to the focal-ration degradation (FRD) in the
fibres, the effective focal ratio at the fibre exit is F/5. The movable (or “deployable”) IFUs
are a unique characteristics of GIRAFFE. These devices can be placed all over the FLAMES
field of view with the exception of the very center of the plate.
Underneath the microlenses, a totally reflecting LLF1 prism sends the light to the fibres. Each
IFU contains twenty associated single fibres and each plate hosts fifteen IFUs. In addition, 15
FLAMES User Manual
VLT-MAN-ESO-13700-2994
24
MEDUSA FIBRE LINK: OVERALL VIEW
In the positioner
Simultaneous
calibration box
Hytrel jacket
1,2 m
Type A: 5 bundles/focal plate
Hytrel jacket
(L: 7,5 m)
1,3 m
1 Fibre per
button
Fork
Fork
1m
5m
Anchored point
Polyurethane
jacket
12 buttons
per sub-slits
13 fibres
per slit
In the positioner
1,2 m
1 Fibre per
button
1,3 m
Type B: 8 bundles/focal plate
Fork
(shrinkable tube)
9 buttons
per sub-slits
Anchored point
(tube ø 4mm)
Polyurethane jacket
(L: 6 m - ø 3 mm)
9 fibres
per slit
Number of buttons :
132 buttons (for star and sky)
Figure 8: Scheme showing the buttons, fibre bundles, and the geometry of the MEDUSA slit.
Sky IFUs are present on each plate. They are identical to the object IFUs, with the exception
that only 1 fibre takes the light from the central microlens.
The fibres of each IFU are organized in a special way on the microlens array, to guarantee
the maximum of contiguity between fibre exit and fibre input. The output of the twenty IFU
fibres plus the Sky IFU fibre constitute an IFU subslit (with therefore twenty-one fibres). In
total, the IFU slit is composed of fifteen subslits. In addition, five subslits contain in addition
a calibration fibre fed by the GIRAFFE simultaneous calibration box.
The center-to-center distance between the fibres in the subslits is only 1.47 times the fibre
diameter core, which implies that the contamination between adjacent fibres is rather high
(about 10%); in any case, a even higher level of contamination is always present at the fibre
entrance level in normal observing conditions.
16.4.1
IFU Orientation
Fibres are kept in constant tension by the springs in the retractors, and the buttons are free to
float in the gripper before been placed. When a button is placed, the fibres is always oriented
along the line which join the position of placement and the retractor (parking position), as seen
in the FPOSS mimic. This means that the IFUs are always oriented (within a few degrees)
with the long side perpendicular to the conjuction fibre-retractor, that is, the long side of the
IFU (3 arcseconds) is perpendicular to the fibre, the short (2 arcseconds) is oriented along the
fibre-retractor direction.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
25
Table 6: Summary of GIRAFFE spectrograph characteristics
Type
Collimator beam
Collimator aperture
HR Echelle
LR Grating
Camera Focal Length
Detector
Scale
Slit height
16.5
Echelle + Order selection Filters
180 mm
F/5
204 × 408 mm, 316 lines/mm, 63.4◦ blaze angle
156 × 204 mm, 600 lines/mm, 38◦ blaze angle
360 mm (mean)
2048 × 4096, 15 µm EEV CCD
0.19 arcsec/pixel
76.8 mm
ARGUS Fibres
The ARGUS system is a fixed array of 14×22 microlenses, similar to the IFUs, located in the
center of Plate 2. ARGUS is further equipped with a focal enlarger system allowing to switch
between a scale of ‘0.5200 /microlens (1:1) to a finer scale of 0.300 /microlens (1:1.67). In addition
to the object fibres, fifteen ARGUS Sky fibres are available on the plate; they are built identical
to the IFU Sky fibres, i.e., with only the central fibre present.
17
GIRAFFE
GIRAFFE1 is a fully dioptric spectrograph with a beamsize of 180 mm and is able to support
the 76.8 mm longslits fed by the different GIRAFFE fibre systems. It has been conceived
to minimize maintenance and night calibrations; special requirements have been introduced
to reduce setup shifts and to obtain accurate re-positioning. In this way, one is able to
use calibrations obtained in the afternoon. Five fibres can be used to obtain simultaneous
calibration spectra and to monitor the instrument drifts. After passing from the slit unit,
the light is sent through the order sorting filters to the collimator. The collimated beam is
dispersed by one of the two high (HR) or low (LR) resolution echelle gratings. After passing
through the collimator again, an intermediate spectrum is produced. Finally the F/2 camera
produces the image on the 2k×4k CCD (see Figure 2.8). A summary of the most relevant
GIRAFFE characteristics is given in Table 6.
The different GIRAFFE sub-units are described in more detail in the following sections.
17.1
Slit Unit
The slit unit contains six slots: five are occupied by the GIRAFFE fibres and one is occupied
by the long slit which can be illuminated by an internal calibration unit. The slit unit is the
most complex mechanical subsystem of GIRAFFE, because it needs a very high stability and
reproducibility. Two movements allow to exchange the fibre slit and to set the fibres in focus.
1
The GIRAFFE spectrograph obtained its name from the first concept for the instrument, in which it was
positioned vertically on the Nasmyth platform.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
26
y, E(PA=0)
x, N(PA=0)
Figure 9: Geometry of the ARGUS microlens array. The arrows • −→ indicate the orientation
of the subslits in the array and the direction of the increasing number of the fibre’s position
in the subslit (PSSN) as given in the static Fibre Binary Table (cf. Section 46). The x and y
coordinate system refers to the X and Y columns of the Fibre Binary Table with which the
ARGUS image array is reconstructed from the fibre’s position in the ARGUS slit (FPS). For
a ARGUS position angle of PA = 0 the North–East orientation on sky is indicated for the
reconstructed image, too. This is the long axis of ARGUS.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
27
In addition, the slit unit is equipped with a number of back-illumination LEDs. These LEDs
are powered and controlled by the Fibre Positioner. They are used to allow the gripper camera
to view the fibre output when positioning. Clearly, since they are lit during the GIRAFFE
exposures, special care was taken in keeping them very well light-tight, in order not to create
light contamination in the spectrograph. Thanks to this system, while one is observing with a
set of fibres on one plate, any set of fibres on the other plate can be prepared by the positioner
for the next observations.
17.2
Filters and the Filter Wheel
After the slit, an interferometric filter selects the light according to the chosen wavelength and
resolution. In addition to excellent transmission and image quality, these filters must fulfill
very stringent requirements on the bandpass edges and blocking over the whole CCD sensitive
bandpass, in order to avoid pollution from adjacent spectral orders. This is very critical,
especially in the blue where the orders are rather short in wavelength. The transmission
curves for all filters can be found in the Appendix.
The thirty GIRAFFE filters are mounted on a filter wheel with four layers, each with twelve
filter positions. A filter is selected by selecting the one of the 4 layers (A-D) and one of the
rotational filter positions (1-12).
17.3
Dioptric Spectrograph
After the light is passed through the filters it is reflected towards the collimator, which works
in double pass, i.e., the light passes through it again after being dispersed. The configuration
angle is six degrees. The main dispersers are two commercial echelles:
The high dispersion grating is a protected silver coated 200 × 400 mm 63.6 degree echelle (R2)
with a high groove density (316 lines/mm), which ensures large orders also in the blue. The
grating can turn on a turntable turret with very high accuracy (0.05 pixel) and repeatability
(0.05 pixel rms). The whole (370 − 950 nm) spectral range is covered by twenty-four setups in
10 grating orders.
The original HR grating was replaced in October 2003 which has lead to an average efficiency
gain of 46 per cent and loss of resolution of 15 per cent.
The low resolution grating has 600 lines/mm and a blaze of 34 degrees. The whole spectral
range is covered with eight setups in 4 orders. The grating size is 150 × 200 mm, which implies
that some vignetting (8 % ) is present due to the geometry of the beam (180 mm). In the
future, the aim is to provide a grating with the same characteristics but a larger size. The
transmission curve of the low resolution grating is also given in the Appendix.
After being dispersed, the light passes again through the collimator, forms a real image at an
intermediate focal plane and is finally imaged by a rather complex F/2 fully dioptric camera.
The camera has seven elements, is thermally compensated for focus displacement through a
system of thermally compensating bars that move one lens inside the camera. Note, however,
that the collimator changes cannot be compensated with this system and that these are instead
compensated through the slit focus movement.
FLAMES User Manual
17.4
VLT-MAN-ESO-13700-2994
28
Scientific CCD “Bruce”
The GIRAFFE EEV CCD (internally named “Bruce”) has an exceptionally good quantum
efficiency in the Blue and UV domain. It is equipped with a continuous flow cryostat, which
provides a high mechanical stability. The liquid nitrogen tank is exchanged every ∼14 days and
a seal between the CCD and the enclosure ensures the thermal insulation of the spectrograph.
Bruce is controlled by a FIERA controller. We have opted not to offer any possibility of
changing the CCD readout characteristics, i.e., windowing, variable read-out speed and binning
are not possible for the user. Note that on-chip binning would undersample some regions of
the spectra also in MEDUSA mode.
A compromise has been adopted for the read-out speed: the chosen setup works at a read
speed of 225 kpixel/sec, which implies a reading time of 42 seconds, and excellent read-out
noise (4.2 e-/pixel at 225 kpix/sec reading speed and low gain). The CCD working temperature is −120◦ C, to minimize some of the blemishes, and to ensure negligible dark current
(1 e− /pixel/hour).
A summary of the GIRAFFE CCD characteristics in the readout mode 225,1×1,low (which
is the only one presently offered) is given in Table 7.
17.5
Spectral Format and Efficiency
In GIRAFFE the spectra are parallel in dispersion along the long side of the detector (i.e., in
readout direction), while on the short side is parallel to the slit. Spectra are curved, with the
central part closer to each other than the edges. The lines of constant wavelength describe arcs
of low curvature with respect to the CCD pixels. This implies that the wavelength coverage
is slightly shifted (by a few Ångstroms) between the fibres at the edge and the fibres at the
center of the slit (which are shifted to the red). An example of a Th-Ar wavelength calibration
spectrum is given in Figure 10, blue is towards the left.
The setups have been fixed to guarantee instrument operability. By turning the grating, different combinations can be obtained for a given central wavelength and grating. The resolving
power and coverage are both function of the grating angle.
For both gratings the spectral orders are quite long with respect to the detector. In order to
guarantee the whole coverage, the grating(s) need to be rotated, which causes the differences
in resolution and spectral coverage between the different setups. A higher angle corresponds to
a higher resolution and to a smaller wavelength coverage. The setups have also been selected
to give a 10% overlap between consecutive setups. This enables an easy connection between
the different spectral slices of the same object.
The higher resolving power of the IFU and ARGUS modes (compared to the MEDUSA mode)
is solely due to the smaller size of the fibres, which projects to 2.4 − 2.6 pixels instead of
3.8 − 4.2 pixels of MEDUSA. The spectral coverage for a given setup is the same in the
MEDUSA, IFU and ARGUS modes.
For a given spectral format, the spectral resolution is very uniform along the chip. Measurements show a variation of the resolving power along the chip of less than 4% rms.
As far as overall efficiency is concerned, the ETC can be used to estimate the instrument
performance in detail. In general, the telescope plus spectrograph delivers a peak efficiency of
FLAMES User Manual
VLT-MAN-ESO-13700-2994
29
Table 7: Measured properties of the GIRAFFE scientific CCD “Bruce” (as from April 2002,
Paranal).
Quantum efficiency
Number of pixels
74.0%
91.5%
89.0%
71.5%
54.3%
26.6%
@
@
@
@
@
@
350 nm
400 nm
500 nm
700 nm
800 nm
900 nm
2048 × 4096
Pixel size
15 µm
Gain (low)
2.25e− /ADU
Readout noise
Saturation (low gain)
Full frame readout (225 kpix/sec)
4.2 e− rms
∼60000 ADU
42 sec
Dark current levels (−120◦ C)
1 e− /pix/h
Fringing amplitude (850 nm)
up to 40%
CTE
> 0.99999
Readout direction
Prescan, Overscan areas
in disp. direction.
Pix. 1-50 and 2098-2148
FLAMES User Manual
VLT-MAN-ESO-13700-2994
30
Figure 10: GIRAFFE CCD image of a Th-Ar calibration lamp. The fibres in the center of
the slit have lines moved towards the blue (left), i.e., fibres corresponding to the center of the
CCD will have a slightly redder wavelength coverage than the ones at the CCD edges.
10%, uniform in the 450 − 700 nm range, which degrades towards the blue (due to optics and
fibre transmission) and towards the red (due to CCD efficiency).
It is important to recall that the efficiency can change substantially within the same setup,
by almost a factor two if the wavelength of interest is close edge of the order. The ETC gives
a very reasonable representation of the final spectral format.
17.6
GIRAFFE Setup Stability and Repeatability
GIRAFFE has been built to be mechanically very stable; the slit exchange mechanics were
designed to ensure high setup repeatability; the basic requirement being that the calibrations
made during the day would well reproduce the night observations. This has been well achieved,
as confirmed form tests in Garching and Paranal. The flexures due to temperature variations
measured in GIRAFFE in long term tests amount to ∼ 0.3 pixels/K along and perpendicular
to dispersion, which translates into typical shifts of < 0.15 pixel in the 12 hours night/day
interval on the detector. The number given above include the repeatability of the setup, since
the tests were done by swapping setups between the tests.
17.7
GIRAFFE Calibration Units
In addition to the calibrations performed through the Positioner calibration Unit (illuminating
all the fibres sequentially, cf. Section 2.3), GIRAFFE is equipped with two calibration units:
Simultaneous Calibration Unit (SCU)
In order to limit the use of time-consuming night calibrations, in each GIRAFFE mode 5 fibres
are devoted to the acquisition of simultaneous Th-Ar spectra illuminated by the SCU during
the science integration. The unit is equipped with a tunable neutral density filter which allows
FLAMES User Manual
VLT-MAN-ESO-13700-2994
31
good Th-Ar exposure levels for integration times between 2 and 120 minutes. If not deselected
in the observing template, these 5 SCU spectra are acquired automatically. Note that for faint
objects the SimCal spectra can contaminate neighbouring spectra (see Sect. 20.1). Users can
choose to take a 60-s exposure with the lamp ON, then the (long integration) science target
with it OFF, then a 60-s exposure with it ON again, in order to eliminate the possibility of
contamination.
Longslit Calibration Unit
One of the GIRAFFE slits is equipped with a longslit unit, fed by a calibration system with
an integrating sphere. The slit is illuminated by an F/5 beam to simulate the F/5 fibre exit.
This slit unit is mainly used for engineering tests and is equipped with one FF, one Ne and
one Th-Ar lamp.
18
UVES – FIBRE mode
In fibre mode, part of the pre-slit area of UVES (see UVES Manual) is substituted by a fibre
projector, which transforms the fibre focal aperture F/3 at the fibre exit into a 25 mm parallel
beam.
In front of the fibre projector a shutter allows the light from the fibres to reach the mode
selection mirror where the fibre mode is selected. After the light passes through a re-imaging
F/10 lens, the regular red UVES slit and the UVES RED arm is used in fibre mode. UVES is
exhaustively described in the UVES User Manual; only a very short description is given here
(see http://www.eso.org/instruments/uves).
The UVES RED arm (λλ 420−1100 nm) is a white-pupil type design. With a beam of 200 mm,
the off-axis parabolic collimator illuminates the echelle grating of 214 × 840 × 125 mm with a
large blaze angle (76◦ ). The echelle is used in quasi-Littrow mode, i.e., the angle of incidence
and the angle of diffraction are equal but in a different plane, which maximizes efficiency.
The grating cross-dispersers provide an order separation larger than 10 arcsec at any wavelength in the spectral range 420 − 1100 nm. This separation allows to host the 8 UVES fibres
(for the 580-nm and 860-nm settings), which have an aperture of 1 arcsecond each and are
separated by 1.47 arcseconds.
The red camera is dioptric (no central obstruction) and provides an external focal plane for
easy detector interfacing and upgrading during the lifetime of the instrument, together with
a large field, good image quality and high optical transmission.
In the red arm, a mosaic of two 4096 × 2048 pixels CCDs is offered, separated by about 1
mm (loss of one order in the gap). The direction of the spectral dispersion (= echelle orders)
is along the larger dimension of the CCDs. The instrument spectral formats (wavelength
coverage, etc.) are always computed for these fixed CCD window settings.
The Arm Selector unit has four positions: Free (direct feed to the red arm), Mirror#1 to
feed the blue arm, Dichroic#1 and Dichroic#2 to feed both arms. In fibre mode, the backside
of Mirror#1 is used to feed the red arm of UVES with the light from the eight FLAMES fibres.
The working position of this unit is determined automatically by the instrument software once
the instrument observing mode is selected.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
32
Table 8: UVES echelle and cross-disperser gratings
g/mm
RED
31.6
g/mm
CD#3
CD#4
18.1
600
312
Echelle gratings
Resolving
Spatial
Power
resolution
2.100.000
0.09”
Blaze
angle
75.04◦
Blaze
Eff. (%)
63
Cross-disperser gratings
Wav. range Average Wav of Peak
Blaze
(nm)
Eff. (%)
Eff. (nm)
Eff. (%)
420 − 680
> 60
520
68
660 − 1100
> 70
770
80
The RED Spectrograph Arm
The red mirror collimator consists of two off-axis parabolas and two flat mirrors. It is of
the “white pupil” type with two 200 mm pupils: one for the echelle and one at the crossdisperser/camera which results in a moderate size of the optical components and a simplified
design.
The red echelle grating is 840 mm long and 210 cm wide. Because grating masters of this
size cannot be ruled, a new process was developed in which a replica is made of two precisely
aligned masters. The result is called a “monolithic mosaic” and has a resolving power on the
order of 2 000 000 and a stable Line Spread Function. The groove density and hence the order
length was selected such that the order length at 990 nm is equal to the CCD length. Further
information on the echelle (and cross-disperser) gratings can be found in Table 8.
The cross-disperser unit is a grating turret with two gratings mounted back-to-back. Selection of the grating is done by rotation of the unit; the angle of the grating is automatically
set according to the required wavelength of the central echelle order. The properties of the
red cross-disperser gratings #3 and #4 can be found in Table 8.
The Camera is dioptric with an external focus to facilitate detector exchange. Focus is set
manually and then maintained automatically by thermal expansion rods in the camera support
structure. The red camera has unvignetted entrance apertures of 230 mm, focal length of
500 mm, and fields of 87 mm diameter. Its image quality is 20 µm on-axis to 30 µm in the
corners (diameter of circle containing 80% of the energy). The transmission curves can be
found in the UVES database available through the instrument ETC.
18.2
Scientific CCD Mosaic “STING”+“NIGEL”
A summary of the properties of the red arm scientific CCDs is given in Tab 9. The detailed
QE curves can be found in the UVES database available through the ETC. The detector in
the red camera consists of a mosaic of one EEV (CCD 44-82) and one MIT-LL (CCID-20)
4k×2k CCD; this to optimize the detector response as a function of wavelength and to reduce
fringing at far-red wavelengths. The gap between the two CCDs is ∼ 0.96 mm. This gap and
the non-perfect alignment of the two chips require a separate extraction of the spectra of the
two chips. The CCD control system (the ESO standard system FIERA) reads the mosaic as
FLAMES User Manual
VLT-MAN-ESO-13700-2994
33
Table 9: Measured properties of UVES-RED scientific CCDs (Dec 99 values).
EEV
89% @ 450nm
89% @ 600 nm
MIT/LL
81% @ 800 nm
46% @ 900 nm
2048 × 4096
2048 × 4096
Pixel size
15 µm
15 µm
Gain (low)
1.6e− /ADU
low: 1.5e− /ADU
3.4 (2.0) e− rms
3.8 (3.4)e− rms
∼65000 ADU
∼40000 ADU
30 sec
30 sec
Dark current levels (−120◦ C)
0.5 e− /pix/h
1.1 e− /pix/h
Fringing amplitude (850 nm)
up to 40%
up to 20%
CTE
> 0.99995
> 0.99995
in disp. dir.
in disp. dir.
Pix. 1-50 and 2098-2148
pix 40-50,2098-3008
< 30µm peak to peak
< 60µm peak to peak
Quantum efficiency
Number of pixels
Readout noise
Saturation (low gain)
Full frame readout (225 kpix/sec)
Readout direction
Prescan, Overscan areas
Flatness
FLAMES User Manual
VLT-MAN-ESO-13700-2994
34
a single image with 100 artificial pixels between the two sensitive areas. The file has to be
split before applying a standard echelle reduction package. Windowing of the CCDs is
not allowed, neither is CCD binning in UVES – FIBRE mode. Only ONE read
out mode of the CCDs is offered in visitor and service mode: Low gain, fast read-out
(225 kpix/sec), 1x1 binning.
The characteristics of these modes are given in Table 9. The linearity of the CCDs is measured to be better than 1% over the range from ∼200 e− to the saturation limit. The CCD
parameters are periodically re-measured at the observatory. The updated values are entered
in the instrument database and are recorded in the FITS headers, for later use in the data
reduction. The cosmetic quality of the scientific CCDs is very good.
The CCD cryostat is attached to the dioptric camera, with the last optical element acting as
the dewar entrance window. The CCDs are operated at a temperature of ∼ 153 K. A liquidnitrogen tank ensures continuous operation without manual intervention for ∼ 2 weeks. The
shutter is located between the cryostat window and the camera. It is actuated by solenoids
with an open/close time of 50 ms. The illumination of the detectors is homogeneous within
50 ms but a minimum exposure time of 0.5 sec is recommended. The reader is referred
to the CCD webpages of the ESO Optical Detector Team for additional general information
on the CCDs and the FIERA CCD Control System (http://www.eso.org/odt/).
18.3
Spectral Resolution and Overall Efficiency
In contrast to slit mode, in fibre mode the user cannot set the resolving power by choosing
the slit width, and the resolving power is determined by the projection of the fibre apertures
on the CCD. The only variable factors which may affect the resolving power are the image
quality of the optics (including the focus and the alignment), CCD effects (chip tilt, diffusion
of photoelectrons, charge transfer), as well as the echelle dispersion. The instrument does not
include a remotely controlled focus adjustment, since the cameras automatically compensate
for temperature variations of the complete optical train from slit to CCD. The image quality
over the entire spectral range is better than 20 − 30 µm over the full CCD (80% of the energy). This implies that, in practice, no noticeable variations are expected between different
exposures. The measured resolving power in Fibre mode is R ≈ 47000.
The efficiency of the RED arm of UVES in fibre mode is about 40% lower than
UVES in slit mode for observations of a single point source.
It is rather difficult, however, to accurately predict the real differences in flux collection between the two modes, because they will be function of the seeing and of how accurate fibre
centering has been performed.
The transmission and reflection efficiency curves of the various optical components and of
the CCDs (cf. UVES database available through the instrument ETC) can be combined to
compute the predicted overall instrument efficiency which turns out to be higher than 10%
from 400 to 800 nm.
FLAMES User Manual
19
VLT-MAN-ESO-13700-2994
35
FLAMES Features and Problems
This section lists all the features and problems common to FLAMES. Several of these topics
are very important for efficient observations with FLAMES.
19.0.1
Maximum reachable S/N ratio
Fibre systems, when coupled to spectrographs, suffer from small photometric instabilities,
whose relevance depends on many factors, including the fibre type, the fibre system design,
the spectrograph design, basically the full path from fibre entrance to detector.
This instability shows up as time variable fringing, additional to the fringing produced by the
CCD. This usually induces limitations in the maximum attainable S/N ratio; the measured
vs. expected S/N ratios depart more and more and the measured S/N ratio tends to a maximum
asymptotic value. For FLAMES, S/N ratios have been obtained in a single exposure up to
S/N of 400. The departure from the photon noise in this regime was very high. We consider
this value as the limiting single exposure S/N value.
19.0.2
Enhanced Dark Current after a FIERA Start-up
When the CCD Control System FIERA has to be restarted, e.g. due to a unrecoverable error
or a general failure of the CCD, the level of the dark current will be higher than the value
measured in the running system; approximately an extra 5 ADU in an hour-long GIRAFFE
exposure following the shutdown and an extra 0.5 ADU RMS noise. It is important to check
the performance of the detectors by taking e.g. a dark exposures of a few minutes in binned
mode. An interval of 2–3 hours is normally sufficient to return to optimal performance of the
CCD.
20
20.1
GIRAFFE Features and Problems
Contamination from Simultaneous Th-Ar Calibrations
Although GIRAFFE has very low level of scattered light, the 5 simultaneous fibres, in particular in the reddest setups may show very strong Argon lines. These lines cannot be suppressed
by any filter, and give visible ghosts (at the level of several ADUs) over a large part of the
CCD area. It appears as a diffuse increase in the background (10 − 20 electons), with an
increase up to 40 − 60 electrons (numbers are indicative) very close to the strong lines. These
ghosts may be very bad for those users interested in faint objects low S/N ratio observations,
since they increase substantially the background. Since the spectrograph is quite stable, the
users who are observing faint objects and who are not interested in accurate radial velocity
determinations should switch the simultaneous calibration OFF. This can be done by filling
in the appropriate field in the FLAMES observing templates. For Medusa mode, the retractor
positions of fibres adjacent to the simultaneous calibrations are as follows; Medusa plate 1;
retractor positions 24, 44, 84, 104, 144, 164, 194, 224, 264. Medusa plate 2; retractor positions
24, 44, 84, 106, 130, 170, 204, 224, 264.
FLAMES User Manual
20.2
VLT-MAN-ESO-13700-2994
36
In-focus Ghosts and Scattered Light
A 3% scattered light level is observed in the reddest 300 pixels of the spectrum; the feature
is rather sharp and most likely caused by a white light ghost, i.e., a reflection inside the
collimator; it affects all the 2048 pixels along the slit direction.
20.3
CCD Defects
The EEV chip has very few cosmetic defects. The most noticeable is a hot column, which
affect all the pixels of row 420 redder than pixel 1270. This column dos not affect the same
spectrum in all setups, due to slight shifts between the different setups. The most affected
spectra are spectrum number 24 in MEDUSA mode, which corresponds to button 58, but a
movement of ±1 spectrum depending on the chosen setup and on the long term spectrograph
spectrograph stability, is possible. Other defects can be generally eliminated by flat fielding.
In the upper red corner (i.e, at fibres with high button numbers) a rectangle of about 350×200
pixels shows a higher level of counts in long exposures (up to about 60 electrons in one hour).
21
21.1
UVES Features and Problems
Fibre Overlap in the 520 nm Setup
Below 500 nm, the order separation becomes too small to accommodate all 8 UVES fibres
without overlap. Therefore, the decker of the UVES slit masks the UVES fibres #8 and #9
(corresponding to buttons 103 and 235, respectively) if the 520 nm setting is selected. If these
two fibres have been assigned to targets, the light will be lost. Hence only 6 fibres are available
in the 520-nm setting.
21.2
Fibre-to-fibre Contamination
Given the limited separation between the UVES fibres, a small degree of contamination exists
between one fibre and its neighbour on the slit. This can be seen easily in Figure 11 showing
a trace of three UVES orders in direction perpendicular to the dispersion: three groups of 8
fibres are seen and the reader can notice that the flux level in the interfibre is higher than
the interorder light and that the flux at the base of one fibre is slightly overlapping with the
neighboring one.
The contamination can be divided into two main components:
1. Diffuse light: this light depends primarily on the total amount of light injected in the
spectrograph; it follows the echelle intensity curve and is estimated at the level of 0.2%
/pixel of the adjacent fibre overall intensity. This implies that if 8 stars of similar
intensity are observed, their overall contribution to diffuse light will be at the 1% level
of a single source; however diffuse light has no spectral features, and it appears as a
’continuous’ source.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
37
Figure 11: This figure shows a trace perpendicular to the dispersion of an UVES-FIBRE
frame, containing three orders. Note the slight flux overlap between contiguous fibres. the
UVES-Fibre data reduction software is designed to deblend the contributions
2. Fibre-to-fibre direct contamination: the wings of two adjacent fibres slightly overlap
and this gives a direct contamination (including spectral features) of one fibre to the
next. This value is however very low, and it increases from 0.13% to 0.5% going from
an integration over 5 to 7 pixels. One fibre has a PSF FWHM of 4.5 pixels (Gaussian
PSF approximation), but note that the PSF is not Gaussian.
This implies that some care should be taken in not placing objects with large differences in
magnitudes and possibly of very different nature (e.g., emission and absorption line objects),
in adjacent positions on the detector. Note that the UVES-Fibre Data Reduction Software
(DRS) has been developed to take into account and eliminate this fibre-to-fibre contamination.
21.3
Spectral Gaps in the RED
The CCD detector in the red arm (see Section 2.4) consists of a mosaic of two chips, separated
by a gap of approximately 0.96 mm. This results in the loss of one echelle order in the recorded
spectrum around the central wavelength selected by the observer. At 580 nm the gap spans
around 5 nm and at 860 nm the gap is 10 nm. The dimension of the gap at any central
wavelength can be predicted with high accuracy (≤ 0.5 nm) using the instrument ETC.
FLAMES User Manual
21.4
VLT-MAN-ESO-13700-2994
38
Optical Ghosts in the far red Spectra
Spectra imaged on the CCD mosaic in the red arm are partly reflected back to the crossdisperser grating through the camera lenses. After a further reflection on the grating, the
second order spectrum is re-imaged by the camera on the CCDs. These “ghosts” appear as
in-focus echelle orders, with a steeper inclination and approximately twice the order separation than the primary spectra. The effect is relevant with the cross-disperser #4 at the far
end of the spectrum (central wavelength 860 nm) where the efficiency of the CCD decreases
and the importance of reflections is higher. On the MIT chip (upper CCD in the mosaic),
reflections from the EEV (lower CCD of the mosaic) are seen. For reference, see the flat-field
data available on the UVES web page. The relative intensity of the ghosts to the primary
echelle orders depends on the shape of the target spectrum. With flat fields and at a central
wavelength of 860 nm, the intensity of the ghost orders is < 1% of the real orders on the EEV
chip and between 1 and 7% on the MIT chip.
21.5
CCD Cosmetic Defects
The CCDs of UVES are of excellent (EEV) or good (MIT/LL) cosmetic quality. The number
of hot or less sensitive pixels is limited (< 0.1%) and has little effect on the quality of the data
because of the good sampling.
The defects which, depending on the signal-to-noise of the spectrum, might be visible in the
extracted data are listed below. In the EEV CCD (blue side of the CCD moisaic) there are
four trails of hot pixels which appear in long exposures (X coordinates 3896, 3963, 4052 and
4140 in an unbinned FITS file, middle of the chip in y). They each affect a single column
(longer dimension of the CCD) and are almost parallel to the echelle orders. They would
appear as broadish emission lines in the bluer part of the extracted spectrum of a faint object.
In the MIT-LL chip (red side of the CCD mosaic), there is a trap in the column X = 1609
which might show up as a slight depression over ∼ 130 pixels in the extracted spectrum of
one order.
22
Preparing the Observations
23
Introduction
Before the actual execution of observations, several steps have to be taken. The preparation
of an observing programme is split in two parts: Phase I and Phase II. In Phase I, i.e., the
application for VLT observing time, the emphasis is put on the scientific justification and on
the technical feasibility of the proposed observations. For the specific case of FLAMES,
the proponents must clearly show that they have (or will have) the proper target
list (including astrometry) prior to Phase II. In Phase II the successful applicants have
to prepare their detailed observing plan including the instrument setups using via the Phase II
Preparation (P2PP) tool.
Prior to Phase II, however, it is fundamental that the applicants have prepared
the proper files containing the target list and have already processed them with
FLAMES User Manual
VLT-MAN-ESO-13700-2994
39
the FPOSS tool.
Together with the general Phase I and Phase II documentation
http://www.eso.org/observing/proposals/
,
the information contained in this chapter (and in Chapter 4) provides guidelines for the Phase I
and Phase II preparation process for FLAMES observations. In Chapter 5 information is given
for astronomers who come to Paranal to observe with FLAMES.
The preparation process can be summarized as follows:
Phase I
• Scientific justification
• Choice of instrument and mode
• Estimate of exposure time to reach the required S/N ratio at the desired resolution.
• Selection of the targets, check availability of accurate coordinates.
• Estimate of telescope and instrument overheads
• Determination of scheduling constraints (e.g. visibility, time-critical observations)
• Overview of observation plan (e.g., target list, calibration needs)
Phase II (only for successful applicants)
(bold represent tasks specific for FLAMES)
• Preparation of the target input files
• Preparation of the positioner Target Setup Files with FPOSS. Users should
pay particular attention to the list of broken fibers (see below Sec. 29)
• Identification of detailed instrument setups
• Preparation of required Observation Blocks.
• Recalculation of exposure time if new version of ETC has been released.
Due to its design and concept, FLAMES is an ideal instrument for observations in service
mode, carried out by the observatory staff (i.e., in absence of the applicant).
All the information necessary to the execution of the observations has to be provided to ESO
in the form of Observation Blocks prepared through the P2PP tool, following the instructions sent to the applicants. The Observatory staff will combine the execution of different
programmes in the same night optimizing the time sequence, seeing, and moon requirements.
Observations carried out with the applicant present at the telescope are referred to as visitor
mode observations. In this mode the astronomer prepares (or finalizes) the OBs at the Observatory in advance of his/her nights. He/she decides about the sequence of observations during
the night, but their execution is, however, still performed by the telescope and instrument
operator.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
40
To facilitate the preparation of Phase I (and Phase II) proposals, in addition to the information
provided in this User Manual, ESO has developed sophisticated Exposure Time Calculators
(ETC), one for GIRAFFE and one for UVES-Fibre (see Section 3.8). The ETC permits one
to estimate the signal-to-noise ratio for a given configuration and exposure time (taking into
account specific atmospheric conditions) and determines the spectral format resulting from
the selected instrument setup.
The Observation Blocks (OB) are prepared using another ESO-provided software tool called
P2PP, see http://www.eso.org/observing/p2pp/.
It is worthwhile recalling that for FLAMES the OB preparation is very simple, while most
of the effort is required to provide objects, fiducial stars and VLT guide stars in the same
astrometric coordinate system with a relative rms accuracy better than 0.3 arcsecs.
A UVES-fibre (since April 2003) and a GIRAFFE (from April 2004) data reduction pipeline
runs at the observatory. It enables automatic extraction and wavelength calibration of all
spectra, in order to check of the quality of the observations (resolution, signal-to-noise ratio
in the extracted spectra). The science data are calibrated with calibration exposures from a
calibration database which is updated when required.
24
FLAMES Modes and basic Choices
After the detailed description of the GIRAFFE and UVES spectrographs, and their subsystems
and functions (Chapter 2), we provide an overview of the different FLAMES observing modes.
FLAMES is equipped with two spectrographs, GIRAFFE and the RED arm of UVES. UVES
can be either used with all 8 fibres acquiring source (or sky) photons, or by using seven fibres
on sources and one fibre to record simultaneously a Th-Ar arclamp spectrum (UVES 7+1).
GIRAFFE can operate in either MEDUSA, IFU or ARGUS mode. Simultaneous observations
with UVES can be carried out with any of the GIRAFFE modes. However, only one GIRAFFE
mode can be used at a time (e.g., it is not possible to observe simultaneously with MEDUSA
+ IFU).
In summary, all the possible modes of FLAMES include:
1. UVES (8 fibres to targets in 580 or 860 nm setup, 1 arcsecond diameter each)
2. UVES (6 fibres to targets in 520 nm setup, 1 arcsecond diameter each)
3. UVES 7+1 (7 fibres to targets + 1 simultaneous calibration fibre illuminated with a
Th-Ar lamp, only in the 580 nm setup)
4. GIRAFFE MEDUSA (131 fibres to targets, 1.2 arcsecond diameter each + 5 simultaneous calibration fibres illuminated with a Th-Ar lamp)
5. GIRAFFE IFU (15 movable rectangular Integral Field Units, 2×3 arcseconds each, made
of an array of 20 fibres + 15 sky fibre units)
6. GIRAFFE ARGUS (Single, fixed Integral Field Unit, consisting of 14×22 microlenses,
with scale of either 0.52 or 0.3 arcsecond each)
FLAMES User Manual
VLT-MAN-ESO-13700-2994
41
7. Any combined simultaneous UVES (or UVES 7+1) plus GIRAFFE mode (two simultaneous GIRAFFE modes are not possible!)
However, in order to insure a manageable calibration database, only a limited amount of setup
combinations (such as CCD setting) are offered.
It is also important to note that in the COMBINED modes the guiding can be performed
only at one given wavelength, even if the two spectrograph have different central wavelength
settings. This could result in some efficiency losses for the cases when the UVES and GIRAFFE central wavelength are far apart and/or when the chromatic atmospheric effects of
the atmosphere are relevant during the exposure (see Section 3.4).
25
GIRAFFE and UVES Standard Settings
To facilitate the preparation of Observation Blocks (Section 3.1), standard settings have
been defined that allow the observer to select a pre-defined instrument setting for which all
parameters are fixed at optimal values and only the exposure time and number of observations
are left to be decided. The observers can only use these standard settings. The
automatic data processing pipelines are available for these standard settings only.
The FLAMES standard settings are given in Chapter 1, and repeated here for the sake of
completeness:
25.1
GIRAFFE Standard Settings
GIRAFFE standard settings are given for the high and low resolution modes in Tables 10 and
11 respectively. Each setting has a unique FITS keyword INS.EXP.MODE, which is the same
as the “p2pp name” given in the tables. In these tables, in addition to the central wavelengths
of the settings, their coverage and resolving power is given, as measured with Th-Ar lines.
Since the coverage varies slightly from fibre to fibre (cf. Section 2.5.5), the coverage given in
the tables is the coverage interval common to all fibres.
FLAMES User Manual
Filt1
1
2
3
4
5
5
6
7
7
8
9
9
10
11
12
13
14
14
15
15
16
17
17
18
19
19
20
20
21
22
22
VLT-MAN-ESO-13700-2994
θ2 Ord p2pp/ETC
61.1
15 H379.0/HR1
58.2
14 H395.8/HR2
63.0
14 H412.4/HR3
59.1
13 H429.7/HR4
55.1
12 H447.1A/HR5A
63.9
13 H447.1B/HR5B
59.1
12 H465.6/HR6
54.5
11 H484.5A/HR7A
63.9
12 H484.5B/HR7B
58.5
11 H504.8/HR8
53.3
10 H525.8A/HR9A
63.2
11 H525.8B/HR9B
57.3
10 H548.8/HR10
62.0
10 H572.8/HR11
55.6
9 H599.3/HR12
60.3
9 H627.3/HR13
52.6
8 H651.5A/HR14A
65.1
9 H651.5B/HR14B
–
8 H665.0/HR15N
56.4
8 H679.7/HR15
61.1
8 H710.5/HR16
51.7
7 H737.0A/HR17A
65.9
8 H737.0B/HR17B
55.4
7 H769.1/HR18
46.9
6 H805.3A/HR19A
60.1
7 H805.3B/HR19B
49.6
6 H836.6A/HR20A
64.9
7 H836.6B/HR20B
53.2
6 H875.7/HR21
43.7
5 H920.5A/HR22A
57.9
6 H920.5B/HR22B
λ3start
370.0
385.4
403.3
418.8
434.0
437.6
453.8
470.0
474.2
491.7
509.5
514.3
533.9
559.7
582.1
612.0
630.8
638.3
647
660.7
693.7
712.9
722.5
746.8
774.5
785.6
807.3
819.5
848.4
881.6
896.0
λ3cent
379.0
395.8
412.4
429.7
447.1
447.1
465.6
484.5
484.5
504.8
525.8
525.8
548.8
572.8
599.3
627.3
651.5
651.5
665.0
679.7
710.5
737.0
737.0
769.1
805.3
805.3
836.6
836.6
875.7
920.5
920.5
42
4
λ3end ∆λ
RMED/(IFU/ARG)
Comments
386.7 16.7
22500 / 36000
404.9 19.5
19600 / 31300
420.1 16.8
24800 / 39000
439.2 20.5
20350 / 32500
458.7 24.7
18470 / 29481
455.2 17.6
26000 / 41500
475.9 22.2
20350 / 32500
497.4 27.4
18529 / 29632
493.2 19.0
26700 / 42700
516.3 24.6
20000 / 32000
540.4 31.0
17750 / 28372
535.6 21.3
25900 / 41400
561.9 28.0
19800 / 31600
584.0 24.3
24200 / 38700
614.6 32.5
18700 / 29900
640.5 28.5
22500 / 36000
670.1 39.4
17740 / 28334
662.6 24.3
28800 / 46000
679 ∼32 ∼17000 / ∼28000 See caption
696.5 35.9
19300 / 30800
725.0 31.3
23900 / 38000
758.7 45.9
17425 / 27869
749.0 26.5
30200 / 48300
788.9 42.0
18400 / 29400
833.5 59.0
13867 / 22175
822.5 36.9
22200 / 35500
863.2 56.0
16036 / 25511
850.9 31.4
28600 / 45500
900.1 51.7
16200 / 25900
956.5 74.9
11642 / 18628
941.9 45.9
19000 / 30400
Table 10: All 31 high resolution setups of GIRAFFE with 316 lines/mm and 63.5 blaze
grating. These setups are valid for observations taken after October 10th 2003 when a new HR
grating was installed. A comparison between the old and new HR gratings is given in Sect. 51.
The “B” settings always have lower efficiency than the “A” settings. In particular, at λcent the
efficiency ratios are approximately as follows; H447.1B/H447.1A=0.6, H484.5B/H484.5A=0.7,
H525.8B/H525.8A=0.7,
H651.5B/H651.5A=0.5,
H737.0B/H737.0A=0.3,
H805.3B/H805.3A=0.7, H836.6B/H836.6A=0.25, H920.5B/H920.5A=0.7. The H665.0
setting covers both Hα and Li 6707. Previous to P74, settings such as H447.1A, H484.5A etc
were called H447.1, H484.5 without the ’A’ suffix.
1
The filter number;
2
The exit angle;
3
The wavelengths at the edges and center of the CCD in nm
FLAMES User Manual
VLT-MAN-ESO-13700-2994
Filter Theta Order p2pp/ETC name
1
32.4
5 L385.7 / LR1
2
27.9
4 L427.2 / LR2
3
32.2
4 L479.7 / LR3
4
26.3
3 L543.1 / LR4
5
30.6
3 L614.2 / LR5
6
34.9
3 L682.2 / LR6
7
24.7
2 L773.4 / LR7
8
29.0
2 L881.7 / LR8
λstart λcenter
362.0 385.7
396.4 427.2
450.1 479.7
501.5 543.1
574.1 614.2
643.8 682.2
710.2 773.4
820.6 881.7
43
λend
∆λ RIFU/ARG RMED
408.1 46.1
12800 8000
456.7 60.3
10200 6400
507.8 57.8
12000 7500
583.1 81.6
9600 6000
652.4 78.3
11800 7400
718.4 74.6
13700 8600
834.3 124.0
8900 5600
940.0 119.0
10400 6500
Table 11: All 8 low resolution setups of GIRAFFE with 600 lines/mm and 34.0 blaze grating.
UVES RED standard settings
Mode
Cross
Disp.
Below
slit filter
RED
RED
RED
CD#3 SHP700
CD#3 SHP700
CD#4 OG590
Min.
Central
Max. NMaxFib Gap
Wav. Wav. (nm) Wav.
(nm)
414
476
660
520
580
860
621
684
1060
6
8
8
1
5
10
Table 12: The 3 UVES red standard settings are listed below. The two CCDs in the red camera
are separated by approximately 0.96 mm; this results in a gap in the wavelength coverage,
approximately centered on the central wavelength. The start and end points of the spectral
ranges reported in the table are generally conservative as they do not include the echelle orders
which are outside the sensitive area of the CCD by more than 50% of their length.
4
Resolving power R for MEDUSA, IFU, ARGUS.
25.2
UVES Standard Settings
The standard settings for UVES are listed in Table 25.2. They are chosen such that together
they cover the optical wavelength domain from 420 − 1100 nm. The wavelength coverage
is computed for the 4k×4k CCD mosaic of the UVES RED arm. The below-slit filters are
used to suppress the second order of the CD gratings or undesired light from entering the
spectrograph. The wavelength coverage is incomplete above 993 nm, due to the absence of
overlap between adjacent orders.
26
Differential Atmospheric Effects
An important problem that cannot be neglected when performing multi-object spectroscopy
in a large field, is the differential refraction of the atmosphere. This is a differential effect in
the sense that the atmospheric refraction index, and hence the direction of propagation of the
light from a given star, changes with both zenith distance and wavelength.
The consequences for astronomical observations are therefore two-fold:
FLAMES User Manual
VLT-MAN-ESO-13700-2994
44
• An achromatic effect since the refraction index at a given wavelength changes – nonlinearly – with the zenith distance. This is very important when observing large fields,
because stars in different position within the field can have significantly different zenith
distances, and therefore do not move in a coherent way across the sky, making it impossible to guide on the whole field.
• A chromatic effect because the refraction index changes with wavelength. As a result,
the red and the blue part of the spectrum do not hit the fibre at the same position, and
therefore part of the stellar spectrum can fall outside the fibre entrance. This effect is
important only when observing over a wide spectral range, especially in the blue; it will
therefore be more relevant for the GIRAFFE low resolution setups and for UVES.
The achromatic effect cannot be compensated, since it is differential across the field and
strongly dependent on the actual zenith distance. It is therefore extremely important to
be aware of its amplitude in different observing conditions, in order to correctly plan the
observations.
A analytical formula that takes into account all the parameters affecting the amount of the
differential atmospheric refraction has been given by Filippenko (1982, PASP 94, 715). According to his calculation, the refraction index, n, of the atmosphere at a certain wavelength,
λ, is given by:
−6
n(λ) − 1 = 10
29498.1
255.4
P [1 + (1.049 − 0.0157T ) × 10−6 P ]
× 64.328 +
+
×
146 − λ−2 41 − λ−2
720.883(1 + 0.003661T )
"
#
where P is the atmospheric pressure in mm Hg (typically 557.25 for Paranal) and T is the temperature in ◦ C (typically 11.5 for Paranal). Once the refraction index is properly determined,
the displacement of the observed astronomical object with respect to its position “without”
the atmosphere is
dr(λ) = n(λ) tan(z),
where z is the zenith distance of the object.
The problem when observing a large field of view is that an object at the field corner will have
a zenith distance z 0 different from that at the field center z, hence its observed position will
be displaced with respect to the “real” one by a different quantity. The size of this difference
is proportional to tan(z) − tan(z 0 ) and therefore varies non-linearly with the zenith distance
of the whole field. Since the telescope is guiding with respect to the center of the field, this
effect causes distortions at the field edges, that change shape with time as the observed field
moves across the sky.
Figure 12 illustrates the size of this effect for an object located 9 arcmin away from the field
center, as a function of the hour angle of the observations and the target declination. Each
curve corresponds to a different declination, and indicates the size of the relative motion
between the center of the field and the object as function of hour angles. The distance
of 9 arcmin has been chosen as the radius enclosing about one half of the field area. For
comparison, Figure 13 shows the same effect for a star located at the edge of the field, i.e.,
at 12.5 arcmin from the field center. The effect is obviously non-linear with the distance from
the field center, becoming rapidly worse towards the edges.
Figures 12 and 13 refer to a central wavelength of 400 nm. Due to the dependence of the
refraction index upon wavelength, the effect would be significantly smaller in the red than in
FLAMES User Manual
VLT-MAN-ESO-13700-2994
45
Figure 12: Distance between the field center and an object located 9 arcmin away from it,
as a function of hour angle and declination of the field. The dashed horizontal lines indicate
the loci of constant 2, 2.5 and 3 airmasses, from bottom to top, respectively. Computed for a
wavelength of 400 nm.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
46
Figure 13: Same as previous figure but for an object located just at the edge of the field (12.5
arcmin from the center). Also for a wavelength of 400 nm.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
47
Figure 14: Relative displacement between the central wavelength λ=400 nm and nearby
wavelengths covered by typical GIRAFFE gratings.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
48
the blue. The central λ of the observations is given as input to the acquisition template, in
order to allow the telescope to guide on the same wavelength. However, different regions of the
spectrum will be displaced with respect to the central one, and for large displacements they
may fall outside the fibre entrance. Figure 14 illustrates the displacement between a central
wavelength of 400 nm and four other wavelengths in a typical GIRAFFE spectrum covering
a range of 60 nm. Two bluer wavelengths (370 and 385 nm) show positive displacement with
respect to the central one (i.e., the offset with respect to their theoretical position is larger)
while two redder wavelengths (415 and 430 nm) show negative displacements.
The FLAMES Fibre Positioner is designed to calculate the mean position of each object
during the exposure, knowing the field coordinates and the time of the observation. As shown
in Figure 12, extremely rapid variations of airmass causes the position of an object close to the
field edge to change by up to 2 arcsec in one hour exposure. For this reason, it is extremely
important to carefully plan the duration of each single exposure, in order to minimize the flux
losses due to the fact that objects far away from the field center may move away from
their fibre in the course of long exposures.
Observers should make sure that their observations are confined to the flat part of the curves
shown in Figure 12. For instance, fields at declinations below −30o can be observed continuously for 3-4 hours before and after their culmination. On the contrary, fields at declination
between 0 and +30o can be observed for no longer than ∼ 1 hour, and as close as possible to
zero hour angle.
In order to allow the Fibre Positioner to calculate the mean position of each object during
the exposure, it is necessary to provide an expected total execution time of the complete
observation defined in the observing block (OB).
27
Preparing the Target Input Files
All the information regarding the targets are usually defined using the Observation Support
Software (OSS), a number of software tools intended to assist the user in this process.
For FLAMES, the OSS consists of FPOSS, the Fibre Positioner OSS, i.e., a software package that takes an input file, with the target coordinate list, and allows the user to define
(automatically and/or interactively) the way the FLAMES fibre have to be allocated to the
targets.
For this reason, FPOSS is basically the core of the preparation of the FLAMES observations.
The subsequent step, the definition of the observing sequence and exposure times with P2PP,
is then relatively straightforward.
The target information flow starts with the creation of the target input file. The latter is fed to
FPOSS, which then generates a Target Setup File containing the target, guide star, observing
mode, fiducial stars fibre, and guide probe assignation. This Target Setup File is associated to
an OB via P2PP. The content of the Target Setup File (plus additional information) is added
as a FITS binary table to the final spectral images.
Since ESO has no means to check the correctness of the input file, the astronomer must be
very careful; an error in such a file will propagate through the whole data flow without being
detected.
The Target Input file is an ASCII file containing the following columns (see Figure 15 for an
FLAMES User Manual
VLT-MAN-ESO-13700-2994
LABEL M67 central field
UTDATE 2001 04 23
CENTRE 08 51 22.82 11 50 09.4
*EQUINOX J2000.0
M67_00005
08 50 15.98
11
M67_00006
08 50 26.13
11
M67_00007
08 51 01.87
11
M67_00012
08 52 04.43
11
M67_00017
08 50 52.99
11
M67_00020
08 50 22.52
11
M67_00022
08 50 33.83
11
M67_00023
08 50 42.31
11
M67_00025
08 51 31.60
11
33
33
34
34
34
34
34
34
34
58.6
59.2
01.5
10.2
16.3
20.8
21.6
22.0
24.2
P
P
P
P
P
P
P
P
P
1
1
1
1
1
1
1
1
1
18.16
19.80
15.27
19.51
17.28
15.53
14.66
18.69
17.48
49
1
2
3
4
5
6
7
8
9
(B−V)= 0.614
(B−V)= 0.034
(B−V)= 0.008
(B−V)= 0.379
(B−V)=−0.008
(B−V)= 0.018
(B−V)=−0.209
(B−V)= 0.013
(B−V)=−0.064
Figure 15: Example of an input file for FPOSS.
example and the FPOSS User Manual for details):
1.
2.
3.
4.
5.
6.
7.
8.
Object ID - MANDATORY Right ascension in hh mm ss.ss (J2000) - MANDATORY Declination in ±dd mm ss.ss (J2000) - MANDATORY Object type: - MANDATORY G = VLT Guide Star (Magnitudes between R∼ 11 and R∼13). Guide stars brighter
or fainter than this will not allow the telescope to apply the AO correction.
F = FACB (astrometric fiducial star: Field Acquisition Coherent Bundle). Due to the
limited dynamical range of the technical CCD, these stars must not have
a difference in R-magnitude exceeding 3 magnitudes. The absolute magnitude
may be from R=8–15.
P = Program target (same as M)
M = MEDUSA target
I = IFU target
J = IFU SKY target
A = ARGUS sky target
U = UVES target
S = Sky (generic, can be allocated by any fibre type)
Priority: 9 to 1, from 9 highest to lower, respectively - MANDATORY Target Magnitude
Target program ID (integer number)
Comments
Fields not labeled as MANDATORY are in fact optional for FPOSS. For possible additional
information contained in the Comment field see the FPOSS manual.
28
Run FPOSS to Prepare the Target Setup Files
The FPOSS manual provides a full explanation of its use, here we give just a brief summary
of the general concepts. The use of the FPOSS follows the following simple steps:
1. Loading of the input file with Guide, Fiducial, and Target stars
2. Selecting the VLT Guide star
FLAMES User Manual
VLT-MAN-ESO-13700-2994
50
3. Selecting the FLAMES observing mode
4. Assignment of the fibres to target objects
5. Assignment of the fibres to sky positions (if needed)
6. Checking configuration over hour angle range
7. Saving of the fibre assignment to the Target Setup File.
Once the saving is performed, a Target Setup File is generated, containing all the information
necessary for the observation, with the exception of the spectrograph setup.
This file is of outmost importance. The files contains a checksum and must not
be edited. If the file is edited, the following process (P2PP) will not proceed and
will not be able to create OBs. Only Target Setup Files created by the FPOSS
are accepted by the system.
29
Broken Fibers
Although FLAMES fibers are mechanically very stable, the gripper might occasionally fail to
move them, leading to the faulty fiber to be disabled. The large majority of the failures is due
to dust on the plate and/or on the button itself and are readily solved (within 1-2 days) by
cleaning the plate and button. However, it might happen that a button remains disable for a
longer time due to a more severe problem.
If the configuration file provided by the user makes use of a broken fiber, the problematic
fiber is simply ignored by the system which places all available fibers leaving the broken
fibers parked. Thus, if during phase II preparation a given target has been allocated to the
problematic fiber, no data will be produced for this target. There is an additional detail. Since
FLAMES has two observing plates available, the fact that a given fiber is disabled in one plate
does not mean that the fiber with the same fiber number is also disabled in the other plate.
For instance, if object XYZ-1 has been allocated to fiber Medusa 116 and this fiber is broken
in plate 1 but not in plate 2, then a spectrum of XYZ-1 will be produced when the user’s OB
is observed with plate 2 whereas with plate 1 no data will be collected.
Since the amount of broken fiber is small (1-2 per plate) most of the objects will have data
produced by both plates. However, if the user has a have-to-observe target, he or she must
pay attention to the list of broken fibers and not allocate this particular (and precious) target
to any of the fibers in the list of broken fibers.
Note that since FPOSS doesn’t know with which plate your configuration will be observed,
it will allow you to allocate broken fibers. It is up to the user to manually correct the
configuration (deleting allocations and re-allocating the target to another fiber by hand as
described in the FPOSS manual) making sure that highly important target are not allocated
to any of the broken fibers2 .
2
The updated list of broken fibers is available at http://www.eso.org/instruments/flames/visitor.html#Fibres
and http://www.eso.org/observing/p2pp/FLAMES/FLAMES-P2PP.html#Fibres
FLAMES User Manual
30
VLT-MAN-ESO-13700-2994
51
Introducing the Observation Blocks
An Observation Block (OB) is a logical unit specifying the telescope, instrument, and
detector parameters and the actions needed to obtain a “single” observation. It is the smallest
“schedulable” entity, which means that the execution of an OB is normally not interrupted as
soon as the target has been acquired. An OB is executed only once; when identical observation
sequences are required (e.g. repeated observations using the same instrument setting, but
different targets), a series of OBs must be constructed. The detailed definition of FLAMES
Observation Blocks and Templates are given in the FLAMES Template Reference Guide [1].
Usually, one OB consists of two separate entities: the acquisition template and the observation
template(s). It is important to recall that, although not mandatory, it is much more convenient
to repeat the fibre positioning when the same objects are observed through different setups,
and/or very long integrations are required, in order to minimize the effects of the atmosphere
(Section 3.4).
P2PP will read the information regarding the targets from the Target Setup File. Note that
only the files produced by FPOSS are accepted by P2PP.
31
GIRAFFE and UVES Exposure Time Calculators
The GIRAFFE and UVES – FIBRE mode Exposure Time and Spectral Format Calculator (ETC) is accessible through the ESO WEB page at
http://www.eso.org/observing/etc/
.
The ETC models the instrument and detector in their different configurations. It can be used
to compute the detailed spectral format (wavelength and order number as function of position
on the detector) and the expected S/N for the specified target, atmospheric conditions as a
function of exposure time. The ETC can also be used to access the efficiency curves of the
various optical components as measured in the laboratory in advance of the installation.
While using the FLAMES ETC, the user has to keep in mind two fundamental points:
1. Some of the transmission factors are mean values: for instance, Table 3 shows how the
corrector transmission varies with the distance from the field center. The ETC assumes
a distance of 8 arcminutes. In the same way, fibre-to-fibre transmission variations are
present at the 5 − 10% level. The ETC values are also mean values.
2. With an aperture of only 1.2 and 1.0 arcseconds on the sky (MEDUSA and UVES,
respectively), the photon-collecting efficiency will strongly depend on the accuracy of the
astrometry. The ETC is set to a default value of 0.3 arcseconds for the average objectfibre displacement. An option allows the user to specify the object-fibre displacement
to evaluate the effects of bad astrometry on the photon-collecting efficiency.
31.1
Choice of the Sample Target
For the input flux distribution to the ETC, four options can be selected: (1) A blackbody
energy distribution at a given temperature; (2) a power-law distribution; (3) a template spectrum: stellar spectra from spectral type O5 to M2, nebular spectra, galaxy spectra, or a
FLAMES User Manual
VLT-MAN-ESO-13700-2994
52
quasar spectrum; (4) a single line at a wavelength, width, and flux level to be specified. In all
cases but point (4) the object magnitude (in a given broad-band filter) has to be entered. For
extended sources, the magnitudes are given per square arcsec. In addition to the target, it is
necessary to enter the sky conditions (phase of the moon and FWHM of seeing disc).
31.2
Choice of Instrument Configuration and Spectral Format
Based on the properties of the optical components of the spectrograph, the format of the
echelle spectrum covered by the detector depends solely on the selected central wavelength.
The instrument templates with standard settings can be selected from the pull-down menu.
The corresponding spectral formats are given in Tables 3.1 and 3.2. The final entry is the
exposure time.
31.3
Exposure Time and predicted Counts and S/N Ratios
The output of the ETC is a table listing the pixel size in wavelength, the computed efficiency,
the total expected electrons for the object and the sky, the maximum pixel intensity (to
monitor saturation), the predicted S/N ratio per extracted pixel in dispersion direction, the
central wavelength and the wavelength bin size.
32
P2PP tool
The Phase 2 PreParation (P2PP) tool allows the observer to construct OBs. An online tutorial for the creation of FLAMES OBs is available at
.
http://www.eso.org/observing/p2pp/tutorials/tut flames.html
If the observations have been prepared carefully using FPOSS to define the FLAMES modes
and targets-to-fibre assignments (as saved in the Target Setup File) and the ETC to define
the required instrument setups and exposure times, the use of the P2PP tool almost trivial:
one acquisition template and one (or several) observing templates have to be combined in one
OB.
32.1
Acquisition Templates
There are 4 acquisition templates for FLAMES. The first three correspond to the three instrument modes (UVES, COMBINED and GIRAFFE). The fourth template is available in
visitor mode only and is the fast acquisition in the ARGUS mode of GIRAFFE only.
In the first three templates the observer has to fill in the same (2) parameters only:
a) Name of the Target Setup File created by FPOSS to be associated to the template (via
a file selector box).
b) Observing wavelength (from a pull-down menu with all FLAMES standard setups). In
case of combined observations, both the GIRAFFE and UVES observing wavelengths
have to be indicated.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
53
Finally, the fourth template is for the fast acquisition mode of ARGUS, where no FPOSS
setup file is needed. See the template reference manual for details.
32.2
Observing Templates
There are 5 observing templates, one for UVES, one for GIRAFFE, one for COMBINED and
two for ARGUS. In the first three cases only a few parameters are required to be given:
a) Setup (grating and central wavelength from pull-down menu)
b) Observing time for each exposure. Note that UVES and GIRAFFE may have different
exposure times!
c) Number of exposures
d) Switch for the simultaneous calibration (GIRAFFE only). Set to OFF in case negative
affects on your observations are expected.
Of course, in case of COMBINED observations, this information needs to be given for UVES
and GIRAFFE separately. For the final two (ARGUS) cases, the OB must include the information above, plus three additional parameters:
e) Number of offsets.
f) List of offsets in Right Ascension.
g) List of offsets in Declination.
32.3
Computing Time Overheads for your Programme
With the Exposure Time Calculator, the user obtains an estimate of the observing time
needed to reach the desired S/N depending on the object magnitude and observing configuration. To compute the total observing time required for the programme, one needs to add the
time for all actions required to carry on the scientific observation. When applying for service
or visitor mode observations, the computation of the overheads is required and has to
be included in the application.
The following estimates of the overheads must be used and are also the basis for the automatic
calculation of execution times within the P2PP tool, used for the final definition of the OBs
in service and visitor mode:
• Target Acquisition: 15 minutes (except ARGUS fast Visitor Mode only which
is 8 minutes)
The target acquisition includes the configuration of UVES fibres, the homing of the
telescope rotator to zero degrees, the swapping of the plates, and the acquisition of the
field: 9 minutes. The telescope preset, acquisition of the guide star, and start of the
active optics account for an additional 6 minutes.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
54
• GIRAFFE and UVES Instrument setup: 1 minute
A new instrument setup takes 1 minutes for GIRAFFE and UVES.
• GIRAFFE and UVES CCD readout: 1 minute
The readout time for the CCD mosaic in the UVES red arm and for GIRAFFE CCD is
1 minute each. In combined mode all CCDs can be read in parallel.
• Plate Configuration: 0 - 20 minutes
Plate configurations take 20 minutes at most (MEDUSA mode). This does not translate
into overheads if the running exposure on the other plate is at least 20 minutes long.
Plate configuration overheads are to be taken into account only when the exposure time
on one plate is shorter than 20 minutes.
• Nighttime Screenflat Calibration: 7 minutes
If attached screen FF calibrations are requested at nighttime, they will need (on average)
7 extra minutes.
• Daytime Calibrations: 0 minutes
Bias frames, flatfield and Th-Ar calibration lamp exposures are taken only during the
day with the same instrument and detector setup as the science exposures. Standard
calibrations are carried out automatically by the Observatory. No overheads need to be
accounted for.
33
The Calibration of FLAMES Data
34
General Concept
Given the possibility of using two spectrographs in many setups, the possibility of obtaining
suitable calibrations has been a constant concern for FLAMES. The operation concept relies
on the fact that all necessary calibrations can be taken during the day, and they have an
accuracy level to guarantee that a sky subtraction to better than 2% is possible. To achieve
this goal a high photometric fibre stability is required, and an overall high instrument setup
repeatability and stability. All calibrations are carried out using the calibration unit of the
Fibre Positioner (see next section and Section 2.3).
With the exception of the attached Nasmyth FF calibrations, the observer is not supposed to
prepare any calibration OBs: calibrations will be provided by the Observatory, following the
FLAMES Calibration Plan [3].
35
Positioner Calibration Unit
The positioner calibration unit has been conceived to provide the user with the following
performance:
FLAMES User Manual
VLT-MAN-ESO-13700-2994
55
• Fibre-to-fibre relative flux illumination flat field: relative illumination better than 0.3%.
This ensures that accurate fibre-to-fibre relative transmission can be derived using the
flat spectra acquired with the positioner.
• Integration time per button: less than ten seconds. For a flat field level of 8000 e− /pixel,
and for a Th-Ar level of at least twenty lines per GIRAFFE setup. This is valid for the
majority of the setups; in the BLUE setups calibrations require longer integration time.
This performance has been obtained using a mixture of hardware and operations. The fibres to
be calibrated are first positioned following a spiral pattern on the plate. In order to guarantee
the same flat field relative illumination, it is necessary to minimize lamp variability. This is
achieved by sweeping the gripper above the fibres several times; the gripper movement is so
accurate to guarantee the same illumination time/fibre to a level of a fraction of percent.
These flats can be therefore used for both, flat fielding and fibre to fibre transmission measurement purposes.
In case of Th-Ar exposures, the gripper moves from one fibre to the next and dwells on the
fibre for a given amount of time, then goes to the next.
36
Nasmyth Screen
To perform a very accurate sky subtraction, it may be useful to acquire Nasmyth screen
flat fields (or so-called attached flats). These flats are obtained by illuminating the closed
Nasmyth shutter with halogen lamps after the observations. The fibres are not moved at all
from their observing positions during this type of calibration. They are maintained with the
same geometry and torsion property. This ensures the minimum difference between observing
and calibration conditions, but on the other hand, especially for wavelengths bluewards of
420 nm, it is very (night-)time consuming. In visitor mode it may be possible to take such
flatfields in the daytime, time permitting.
37
Simultaneous Calibrations
GIRAFFE is equipped with five simultaneous calibration fibres per slit. Unless explicitely
avoided by the user in the observing template, every spectrum contains five simultaneous arc
spectra, evenly located along the CCD 2k width. These spectra can be used to track the
wavelength solution for all the fibres.
Tests on solar spectra during GIRAFFE integration in Garching reached high accuracy over
a few days, and tests on stars during commissioning showed that an accuracy of 70 m/sec can
be obtained on a time basis of a few hours for slowly rotating cool stars. This was reached
in 15 minute exposures for objects brighter than 14.3 magnitude in the H9 setup (cf. The
ESO Messenger, 110, 1). More accurate and detailed long term RV studies are presently being
carried out on old open clusters.
For FLAMES-UVES the radial velocity instrumental error is about 20 m/s when using the 7+1
mode which includes a fibre dedicated to simultaneous calibration (580-nm only, Astronomy
& Astrophysics, 421, L13). Investigations are underway in order to try and reduce this error
to the 10 m/s level.
FLAMES User Manual
38
38.1
VLT-MAN-ESO-13700-2994
56
Longslit Calibrations
GIRAFFE Longslit Unit
One of the slits of GIRAFFE is equipped with a longslit unit, which is used for engineering
purposes; calibrations obtained with this unit are of no interest for the observer.
38.2
The UVES Calibration Unit
The UVES calibration unit is a mechanical structure mounted on the Nasmyth rotator flange,
which in the case of UVES is kept fixed during observations. It hosts continuum lamps which,
in combination with various filters, are used for flatfield calibration and one Th-Ar lamp for
wavelength calibration. The lamps are mounted on an integrating sphere and relay optics
simulate the F/15 telescope beam. The light from the lamps is fed into the instrument beam
by 45◦ mirrors mounted on a slide.
The flatfield spectra provide a good correction for the blaze function of the echelle. They
correct the pixel-to-pixel variation in CCD sensitivity as a function of the wavelength of
the light. In the red part of the spectra (λ ≥ 650 nm) narrow fringes with peak-to-valley
amplitudes up to 30% are present on the EEV CCD of the mosaic. On the MIT-LL CCD, the
fringes are less sharp and of smaller amplitude.
In the fibre mode of UVES, long slit spectra are acquired with a slit longer than the extent
of the fibre slit to ensure that even in case of small shifts between the observation and the
calibrations, the detector area covered by the fibres is covered by the long slit spectra. Pixelto-pixel variations can be eliminated in this way, up to a S/N ratio of TBD.
39
Fibre to Fibre Transmission (Sky Subtraction)
When dealing with fibre spectra proper sky subtraction is a concern. In the present scheme
(i.e., without nod and shuffle technique) it is critical to characterize the fibre to fibre relative
transmission with excellent accuracy. This task is not always trivial, because fibres may
develop photometric instabilities which depend on their history and on the way they are
routed. In FLAMES we have ensured that in normal operating conditions, the fibre system is
constant to better than 1% stability. At this point, the most critical issue is to find a way to
uniformly illuminate the fibres. This task is done by the positioner.
Other steps to obtain a correct sky subtraction involve a) the knowledge of the transmission of
the corrector, which is given in Table 2.1 as a function of wavelength and position on the field
of view and b) a good spatial distribution of the fibres, which can account for sky variations
in the field of view. It is also important that enough fibres are allocated to the recording of
the sky.
FLAMES User Manual
40
40.1
VLT-MAN-ESO-13700-2994
57
Special Calibrations
Detector Flats
Detector flats from direct illumination of the CCD through the camera only are taken at regular
intervals by the maintenance staff to monitor the CCD performance. They (and all other
calibrations) are available from the ESO VLT Science Archive at http://archive.eso.org.
40.2
Use of Telluric Standard Stars to correct for Fringing or atmospheric Lines
Stars with featureless spectra (typically white dwarfs or fast rotating hot stars) can be used to
provide a good template to correct for fringing as an alternative to the use of internal flatfield
lamps. These spectra can also be used to identify and estimate the depth of atmospheric H2 O
and O2 absorption lines. In case these are required, we suggest that the users insert some of
these objects among their targets.
41
FLAMES Science Calibration Plan
Table 13 provides a summary of the FLAMES Science Calibration Plan as defined in the
FLAMES Calibration Plan [3]. We note that specphot standards are generally selected by the
nighttime astronomer from the list available at http://www.eso.org/observing/standards/spectra/.
42
FLAMES Observing Operations
This chapter explains in some detail the operation of FLAMES. A sound understanding of
the rather peculiar FLAMES operations procedure is crucial for good planning of visitor AND
service mode observations. We summarize again a number of particulars which need to be
taken into consideration for FLAMES operations:
1. The relative distance between objects is going to change during an observation, therefore
in general long observations should be avoided.
2. Observations of the same objects at different wavelengths or multiple observations of the
same objects should be executed after re-positioning of the fibres only.
3. The Fibre Positioner configures while observing; this implies that two OBs are running at
the same time. Also, the positioner needs to know the mean time of the next observation
while the current one is still running. In practice, it gives some rigidity to the whole
operation scheme.
4. When used in combined mode, FLAMES produces UVES and GIRAFFE frames.
5. Each FLAMES sub-system (Positioner, GIRAFFE, UVES) has its own Observing Software (OS). The complete system is coordinated by the FLAMES Super-OS which is the
FLAMES User Manual
VLT-MAN-ESO-13700-2994
58
FLAMES – UVES Science Data Calibration Plan
(per instrument setting, i.e. plate, fibre mode, and central wavelength)
Calibration
Fibre Flatfields
number
3
frequency [1/days]
1/1
Slit Flatfields
attached Fibre Flatfields
Wavelength
3
n
1
1/7
o.r.
1/1
Sim. Fibre Order Definition
Sim. Fibre Format Check
Bias
Dark
1
1
5
3
1
1
1
1
/
/
/
/
1
1
7
30
purpose
pixel-to-pixel sensitivity variations
fibre-to-fibre transmission
fibre localization
fibre PSF modeling
blaze correction
pixel-to-pixel sensitivity variations
high-precision flatfielding
dispersion solution
resolving power
order and background definition
dispersion guess solution
master biases, bias characteristics
master darks, dark current, cosmics rate
FLAMES – GIRAFFE Science Data Calibration Plan
(per instrument setting, i.e., plate, fibre mode, resolution and central wavelength)
Calibration
Flatfields
number
1
frequency [1/days]
1/1
attached Flatfields
Wavelength
n
1
o.r.
1/1
Bias
Dark
IFU: Flux Standard
+ attached Flats
ARGUS: Flux Standard
+ attached Flats
5
3
n
n
1
3
1/7
1 / 30
o.r.
o.r.
1/1
1/1
purpose
pixel-to-pixel sensitivity variations
fibre-to-fibre transmission
fibre(=spectra) localization
high-precision flatfielding
dispersion solution
resolving power
slit geometry
master biases, bias characteristics
master darks, dark current, cosmics rate
response correction, flux calibration
rel. trans. IFU array / Sky fibres
response correction, flux calibration
rel. trans. ARGUS array / Sky fibres
o.r. = on request only, corresponding OBs to be provided by user
n = number to be defined by user
Table 13: Summary of FLAMES Science Calibration Plan
only OS allowed to talk to the Telescope Control Software (TCS). The frames produced
by the spectrographs are complemented with information coming from the TCS.
FLAMES User Manual
43
VLT-MAN-ESO-13700-2994
59
During the Night
Observations with FLAMES are carried out at the Console of UT2, located in the VLT
Control Building just below the Paranal summit. From there, all telescopes and instruments
are remotely controlled. The telescope and instrument operator carries out the observations
and checks that they perform correctly; the main responsibility of the visiting astronomers is
the selection of the OBs, based on the sky conditions and on the results of the first observations.
The GIRAFFE and UVES – FIBRE mode raw data are saved in the FLAMES workstation.
After the data has been transferred to the Archive workstation, copies of the files are received
on the astronomer’s offline workstation and on the pipeline workstation where the automatic
data reduction is running. The pipeline products are eventually forwarded to the astronomer’s
offline workstation (Note that the UVES – FIBRE pipeline has been available from April 2003,
the GIRAFFE pipeline from April 2004).
Given the necessity to run target assignation at the Fibre Positioner together with target
observation at the telescope, two BOBs (Broker for Observation Blocks) are running simultaneously. Although the two BOBs are perfectly symmetric and exchangeable, for the sake of
simplicity we will call BOB-obs the one observing and BOB-config the one configuring.
The Target Setup File generated by the FPOSS (associating the fibres to the object coordinates) is linked to the OB through P2PP.
The OB is read by BOB-config and the acquisition template is executed. The positioner
SW knows the actual time, the expected execution time of the OB (as provided by the user)
and computes the mid-time of the new observation. The coordinates are transformed into
plate R, θ positions, the back-illumination is switched on (about 30 milliseconds/fibre) and
the buttons are placed. Once the configuration of the plate is completed, it receives an unique
identifier (say plate-1) and a validity time-stamp is generated.
When the OB is re-run (on BOB-obs), it looks for the status of plate-1. Finding it at the
configuration position, i.e., at the robot, and knowing that no other observation is running on
the telescope side, it sets the telescope rotator back to the home (=zero) position, disengages
the current plate, rotates the tumbler, and engages the new plate.
At this point the center field coordinates are sent to the telescope and to the rotator. The
VLT field acquisition can start: search for guide star, closing of the active optics loop, and
field stabilization. The four FACBs check that the fiducial or reference stars are indeed in
the right position or of needed telescope offsets are computed and applied. The acquisition
template is now finished and the observing template can start.
While running the observing template on BOB-obs, the next OB can be selected and run on
BOB-config which, after some sanity checks, starts the configuration of the next field.
Clearly this cycle is very critical, and once started, there is not much room for manoevre.
Usually, if some observations require repetition, the plate will need to be reconfigured, the
field be re-acquired and re-executed after another OB has been executed.
43.1
Pointing and Guiding
FLAMES is not equipped with any auxiliary slit viewer or imaging system in addition to the
4 FACBs, therefore the whole system relies on the (relative) accuracy between the targets, the
FLAMES User Manual
VLT-MAN-ESO-13700-2994
60
VLT guide star and the 4 FACBs.
After the telescope has been pointed, the images from the 4 FACBs are recorded on the technical CCD and the centroids in each of the quadrants are computed and offsets calculated. The
fiducial stars in the FACBs are therefore the ONLY sources linking the geometry
of the plate to the sky.
The target, VLT guide star and fiducial (FACBs) coordinates must be in the
same reference system (i.e., their coordinates must be computed from the same
astrometric solution) with a relative accuracy better than 0.3 arcseconds to avoid
wasting telescope time.
To help the users in their observations preparation, ESO has performed a pre-FLAMES stellar
survey using the Wide Field Imager (WFI) at the ESO-MPI 2.2-m telescope:
http://www.eso.org/science/eis
.
In order to guarantee a correct centering and offset calculation, FACB reference stars should
not have close visual companions (within 3 arcseconds). Telescope Guide Stars should
have magnitude R between 11 and 13. FACB stars should be brighter than R∼
15 and be of comparable magnitude due to the limited dynamic range (0–4096)
of the technical CCD (3 magnitude range at most)
Recent images will help to minimize errors due to neglected proper motion in the targets /
guide / fiducial stars. Note also that the direct use of the DSS coordinates is not accurate
enough (Momany et al. 2000), unless proper motions are taken into account.
The tracking of the telescope is corrected for errors of low frequency (< 1 Hz) by an autoguiding
facility. The autoguider makes use of a guide star observed by the guide probe in the adaptorrotator that is moved into the telescope beam. The guide star is selected by the observer in
the input file to the FPOSS and needs to be in the same coordinate system as the targets and
of the fiducial (FACBs) stars.
Once the telescope acquisition and active optics correction is executed, some small shifts may
still be present between the telescope and the target coordinates. When the four FACBs start
working, the offsets of the 4 stars are computed and the operator may apply them to center
fiducial stars on the FACBs. After the centering is considering satisfactory, the observing
template can start and the science integration proceeds.
43.1.1
ARGUS fast observations
In Visitor mode only, it is possible to move from field to field and take observations of different
science targets at different wavelengths without reconfiguring plate 2, at the centre of which
ARGUS is located. These observations rely on the VLT guide star, and hence do not use the
FACBs. These observations are performed using the FLAMES giraf acq argfast acquisiton
template. The 15 ARGUS sky single fibres are placed in a circle with radius defined by the
user. Use of the template saves time as swapping the plates back and forth is not necessary.
However, swapping is avoided only if (a) The ARGUS sky fibres are at the same radius and/or
not used (b) The plate scale is the same for the two observations.
Note that the FLAMES giraf acq argfast template can be used with either FLAMES giraf obs argoff
(science target with offsets), FLAMES giraf obs argstd (standard star observation with offsets) or FLAMES giraf obs exp observation templates.
FLAMES User Manual
43.2
VLT-MAN-ESO-13700-2994
61
Evaluation of the Results, Offline Data Analysis
At the end of each integration, the CCD frames are read out by the FIERA controller and
transferred to the instrument workstation and subsequently to the archive.
During the readout the frames are displayed automatically on two Real-Time Display (RTD)
panels (one for GIRAFFE, one for UVES) for first inspection using the standard RTD tools.
More detailed analysis of the new exposures or previous exposures has to be carried out on
the astronomer’s offline workstation where copies of the raw files are available within a few
seconds. After pipeline processing, the pipeline products are also forwarded to the offline
workstation from where they can be accessed and inspected by the astronomer. Standard
data reduction tools like MIDAS, IRAF, or IDL are available for this purpose.
The rather complex data structure of FLAMES raw frames is described in the next section.
44
FLAMES Raw Data Structure
Given the complexity of the instrument, the data must carry all the relevant information about
the objects and the instrument configurations through the whole data-flow process. In the case
of FLAMES, two binary tables associated to the detector image are of outmost importance,
because they contain all the information to associate the spectra to the objects.
Both GIRAFFE and UVES – FIBRE mode FITS data have the same data and header structure
and include 3 FITS HDUs (Header Units):
HDU1: image the image header contains in addition to the primary FITS keywords additional keywords for the status of telescope, positioner, spectrograph, detector, templates,
etc. during the exposure.
HDU2: OzPoz table associates the objects to the fibre buttons. This is basically the same
table as was provided in the Target Setup File for FPOSS, plus additional information
from OzPoz, such as the R and θ position of each button on the plate and the corresponding positioning errors. The table header contains keywords related to the fibre
positioning process, e.g., the time and wavelength for which the field was configure.
HDU3: FLAMES FIBRE Table contains the fibre description: association between fibre
buttons and position in the subslit and slit, measured (laboratory) fibre transmission at
different wavelengths, fibre bundle. For the ARGUS fibre bundle, the X and Y position
of the individual fibre in the reconstructed image matrix is given.
45
HDU2: OzPoz table
The OzPoz binary table will be different for every frame; this table associates the objects to
the fibre buttons. The basic information for this table is taken from the Target Setup File
(association object-to-fibre and object characteristics). This information is complemented by
OzPoz with all information related to the positioning of the fibres.
The table is structured as:
FLAMES User Manual
Col 1: Object
Col 2: RA
62
Identification (from Target Setup File, column 1)
Right Ascension (from Target Setup File, column 2)
Col 3: DEC
Col 4: R
VLT-MAN-ESO-13700-2994
Declination (from Target Setup File, column 3)
Button R position on plate (microns)
Col 5: R Error
Col 6: Theta
Error in R (microns)
Button θ position on plate (radians)
Col 7: Theta Error
Col 8: Type
Error in θ (microns)
Object type (MEDUSA, IFU, etc. . .)
Col 9: Button
Col 10: Priority
OzPoz button number
Object Priority (from Target Setup File, column 5)
Col 11: Orient
Button Orientation
Col 12: In Tol
T or F if positioned or not within tolerance (40 microns = 0.08”)
Col 13: Magnitude
Target Magnitude (from Target Setup File, column 6)
Col 14: Comments
User comments (from Target Setup File, column 8)
The table FITS header contains additional information from the configuration process. The
following compiles the most important keywords:
Keyword
Example Value
Comment
=================================================================================
--- Information from Target Setup File -----------------------------------------FILENAME
CENRA
CENDEC
CENEQNX
ALLOCGUI
ALLOCOBJ
ALLOCSKY
’w_Cen_.COMMED8.025151’
201.700124999997
-47.5219444444436
2000.
4
122
18
Configuration file name
13:26:48.03, Field centre mean RA (Degrees)
-47:31:20.0, Field centre mean Dec (Degrees)
Equinox of Field Centre (FK5 Julian)
Number of allocated FACB stars
Number of allocated objects
Number of allocated sky positions
--- Information from Configuration Process -------------------------------------ACTMJD
ACTUTC
ATMPRES
ATMRHUM
ATMTEMP
PLATE
FACBWLEN
GIRAWLEN
UVESWLEN
52808.9479166665
’2003-06-18T22:45:00’
745.7
8.
13.4
1
730.
679.7
580.
Actual MJD of tweak time (*)
Actual UTC of tweak time (*)
Atmospheric pressure (millibars)
Atmospheric relative humidity (percent)
Atmospheric temperature (celsius)
Identifier of the used positioner plate
FACB wavelength (nm)
GIRAFFE wavelength (nm)
UVES wavelength (nm)
-- If ARGUS was used ----------------------------------------------------------ARGSUSED
ARGSCALE
T
’1:1’
Flag indicating if ARGUS used
ARGUS Scale
FLAMES User Manual
ARGPOSAN
ARGANGLE
90.
0.
VLT-MAN-ESO-13700-2994
63
Position Angle of ARGUS on sky (Degrees)
Orientation of ARGUS (Degrees)
================================================================================
(*) tweak time refers to the time for which the field was configured.
46
HDU3: FLAMES FIBRE Table
This table is a ’static’ table, in the sense that for all files of a given period it should be
the same. This table contains the association between the fibre buttons and retractors and
the slit position, and the transmission of the fibres as measured in the laboratory before
shipping to Chile. The table is changed only when some major problem or change occurs,
e.g. that a fibre subsystem needs to be replaced with a spare. Several of the data contained are useful for engineering purposes, the reader is invited to concentrate on field 2
(FPS) and field 8 (RP), which give the position in the slit and the position number of the
button on OzPoz respectively. One example of this table can be retrieved via web from
http://www.eso.org/observing/dfo/quality/GIRAFFE/txt/fibre.dat
.
Schematic examples for the raw data format of the different fibre types are given in Appendix 48.
Description of the table columns:
Col 1: Slit
Slit name from fibre type and plate number
Col 2: FPS
Progressive fibre position number in the slit
Col 3: SSN
Subslit number
Col 4: PSSN
Fibre Position in the subslit
Col 5: Retractor
Col 6: BN
Col 7: FBN
Serial Number of the retractor
Serial Number of the button used in the retractor
Serial Number of the fibre used for the button
Col 8: RP Retractor position on the plate. This number corresponds to the fibre number
used e.g. in FPOSS. All even numbers are MEDUSA fibres
Col 9-17: wave Fibre Transmission values as measured in the lab. Each column is a different wavelength
Col 18: X
x position of fibre in the reconstructed image matrix
Col 19: Y
y position of fibre in the reconstructed image matrix
Col 20: FPD Fibre Position on the Detector. For all setups except ARGUS this is the
same as FPS. For ARGUS it is reversed. Added April 2004.
Note that the ARGUS image reconstruction using the x and y columns for the table will give
the image in the standard North – East orientation on sky. If the ARGUS position angle was
0 (ARGPOSAN = 0), N is along the x axis and E along the y axis (cf. Figure 16). The position
angle is counted in the standard sense, i.e., N to E.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
64
E
N
Figure 16: Image reconstruction of the ARGUS integral field unit using the x, y columns
given in the FLAMES Fibre Table. The spectrum was taken with ARGUS at a position angle
0. The corresponding sky orientation is indicated. The image axes are in arcseconds using a
scale of 0.52 arcsec/microlens corresponding to the ARGUS scale 1:1.
47
Appendix
48
FLAMES Raw Data Spectral Format
The following figures give a schematic view of the spectral format for the different FLAMES
fibre types on the raw images.
FLAMES User Manual
48.1
VLT-MAN-ESO-13700-2994
65
GIRAFFE - MEDUSA
Figure 17: Schematic layout of the MEDUSA spectral format: blue solid lines: object fibres,
red dots: calibration fibres. The direction of the increasing fibre number in the slit (FPS) and
increasing wavelength λ are indicated.
FLAMES User Manual
48.2
VLT-MAN-ESO-13700-2994
66
GIRAFFE - IFU
IFU configuration shown for PA=0 deg.
East
4
Y
PSSN (1−20)
20
11
10
1
3
18
19
12
9
2
3
2
17
16
13
8
5
4
15
14
7
6
2
3
1
1
4
5
Plate
North
IFU
Fiber
Retractor
6
X
Notes: 1) Position Angle PA = 315 deg − ORIENT in binary OzPoz table. PA=North−East.
2) For IFUs with SKY fibers, the PSSN numbers should be increased by 1.
3) X,Y and PSSN can be found in the binary FLAMES FIBER table.
Retractor number
07
25
43
59
77
95
113 131 147
165
183 203
219
237 253
Figure 18: Top panel: Geomtric layout of IFU fibers, including the X,Y and Position of
the sub slits for an individual IFU. The IFU orientation (long/short axis) on the plate is also
shown. Bottom panel: Schematic layout of the IFU spectral format (not to scale): blue solid
lines: object fibres, red dots: calibration fibres, green dots: sky fibres. The direction of the
increasing fibre number in the slit (FPS) and increasing wavelength λ are indicated, as well
as the retractor number for each IFU.
FLAMES User Manual
48.3
VLT-MAN-ESO-13700-2994
67
GIRAFFE - ARGUS
ARGUS
SSN:
15 14
FPS
PSSN
Note that SSN, FPS and PSSN increase from RIGHT TO LEFT
in the current version of the fibre table.
13 12
11 10
9
8
7
6
5
4
3
2
1
4000
y
(pixels)
3000
Lambda
2000
1000
0
0
1000
x (pixels)
2000
Figure 19: Schematic layout of the ARGUS spectral format: blue solid lines: object fibres,
red dots: calibration fibres, green dots: sky fibres. The direction of the increasing fibre number
in the slit (FPS) and increasing wavelength λ are indicated. Note that the directions of FPS,
SSN and PSSN are inverted w.r.t. MEDUSA and IFU.
FLAMES User Manual
48.4
VLT-MAN-ESO-13700-2994
68
UVES - FIBRE
Figure 20: Schematic layout of the UVES – FIBRE spectral format for one order: blue solid
lines: object fibres, red dots: calibration fibre. The direction of the increasing fibre number
in the slit (FPS) and increasing wavelength λ are indicated. Redder echelle orders are to the
left.
FLAMES User Manual
49
VLT-MAN-ESO-13700-2994
69
Characteristics of GIRAFFE Filters
Figure 21: Transmission of the GIRAFFE High Resolution filters 01 − 06. Wavelength is in
nm.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
70
Figure 22: Transmission of the GIRAFFE High Resolution filters 07 − 12. Wavelength is in
nm.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
71
Figure 23: Transmission of the GIRAFFE High Resolution filters 13 − 18. Wavelength is in
nm.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
72
Figure 24: Transmission of the GIRAFFE High Resolution filters 19 − 22 and Low Resolution
Filters 01 − 02. Wavelength is in nm.
FLAMES User Manual
VLT-MAN-ESO-13700-2994
73
Figure 25: Transmission of the GIRAFFE Low Resolution filters 03 − 08. Wavelength is in
nm.
FLAMES User Manual
50
VLT-MAN-ESO-13700-2994
FLAMES calibration times
74
FLAMES User Manual
51
VLT-MAN-ESO-13700-2994
75
Comparison between old and new HR gratings
In early October 2003, the high resolution grating on GIRAFFE was changed, leading to an
increase in sensitivity at the loss of spectral resolutions in some setups. A comparison between
old and new gratings is shown in the following table.
1
The filter number;
2
The central wavelength in nm;
3
Resolving power R in MEDUSA mode for old and new HR gratings;
4
Average efficiency for old and new HR gratings in percent.
FLAMES User Manual
Mode
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
Medusa
UVES 6FIB
UVES 7+1/8FIB
UVES 8FIB
VLT-MAN-ESO-13700-2994
76
λ(nm) ETC name t(arc,robot) t(FF,robot) t(FF,screen)
379.0
HR1
600
1250.0
5000
395.8
HR2
600
200.0
600
412.4
HR3
600
235.3
600
429.7
HR4
600
83.0
200
447.1A
HR5A
600
80.0
150
447.1B
HR5B
600
80.0
300
465.6
HR6
600
26.3
100
484.5A
HR7A
600
30.3
50
484.5B
HR7B
600
30.3
140
504.8
HR8
600
12.5
30
525.8A
HR9A
600
16.7
60
525.8B
HR9B
600
16.7
80
548.8
HR10
600
5.3
25
572.8
HR11
600
6.7
30
599.3
HR12
600
3.2
15
627.3
HR13
600
3.8
20
651.5A
HR14A
600
6.1
11
651.5B
HR14B
600
6.1
25
665.0
HR15N
600
6.1
10
679.7
HR15
600
3.2
10
710.5
HR16
600
3.8
10
737.0A
HR17A
600
10.0
6
737.0B
HR17B
600
10.0
22
769.1
HR18
600
6.2
7
805.3A
HR19A
600
8.7
5
805.3B
HR19B
600
8.7
7
836.6A
HR20A
600
25.0
4
836.6B
HR20B
600
25.0
20
875.7
HR21
600
11.1
4
920.5
HR22A
600
21.5
6
920.5
HR22B
600
21.5
10
385.7
LR1
600
133.3
300
427.2
LR2
600
13.3
60
479.7
LR3
600
8.0
40
543.1
LR4
600
1.7
12
614.2
LR5
600
1.7
8
682.2
LR6
600
1.4
6
773.4
LR7
600
2.1
3
881.7
LR8
600
5.6
3
520
–
80
30
80
580
–
60
20
40
860
–
60
20
40
Table 14: Integration times in seconds for ThAr arcs and W flats, for both Robot calibrations
and attached Screen flats. IFU/Argus times are twice the Medusa values. Configuration time
is excluded.
FLAMES User Manual
Filter1
λ2center
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
379.0
395.8
412.4
429.7
447.1
465.6
484.5
504.8
525.8
548.8
572.8
599.3
627.3
651.5
679.7
710.5
737.0
769.1
805.3
836.6
875.7
920.5
VLT-MAN-ESO-13700-2994
77
Order Order
∆λ
∆λ
R3
R3 AvEff %4 AvEff %4
(old) (new) (old) (new) (old) (new)
(old)
(new)
15
15 16.7
16.7 22500 22500
29.0
43.5
14
14 19.5
19.5 19600 19600
26.7
50.3
14
14 16.8
16.8 24800 24800
30.3
36.9
13
13 20.5
20.5 20350 20350
36.4
55.7
13
12 17.6
24.7 26000 18470
29.2
48.3
12
12 22.1
22.2 20350 20350
41.8
60.9
12
11 19.0
27.4 26700 18529
32.7
45.6
11
11 24.6
24.6 20000 20000
44.2
57.1
11
10 21.3
31.0 25900 17750
38.4
39.0
10
10 28.0
28.0 19800 19800
44.4
58.4
10
10 24.3
24.3 24200 24200
43.4
41.6
9
9 32.5
32.5 18700 18700
38.2
58.4
9
9 28.6
28.5 22500 22500
47.6
51.9
9
8 24.3
39.4 28800 17740
28.7
54.2
8
8 35.9
35.9 19300 19300
44.5
61.3
8
8 31.3
31.3 23900 23900
42.0
44.1
8
7 26.5
45.9 30200 17425
25.6
64.8
7
7 42.0
42.0 18400 18400
46.2
65.4
7
6 36.9
59.0 22200 13867
42.0
50.0
7
6 31.4
56.0 28600 16036
29.8
70.6
6
6 51.7
51.7 16200 16200
42.9
62.6
6
5 45.9
74.9 19000 11642
50.7
51.7
Table 15: Comparison between the GIRAFFE HR grating used pre and post-October 10 2003
FLAMES User Manual
VLT-MAN-ESO-13700-2994
—— oOo ——
78